CORRELATION OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES 251
GEOLOGICA CARPATHICA, 55, 3, BRATISLAVA, JUNE 2004
251260
LITHOLOGICAL AND MAGNETOSTRATIGRAPHIC CORRELATION
OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES
(SE BULGARIA)
STILIYAN MOSKOVSKI
1
, VASSIL KARLOUKOVSKI
2
, ZLATKA MILAKOVSKA
3
,
ALEXANDRA HARKOVSKA
3
and MALCOLM PRINGLE
4
1
Faculty of Geology and Geography, Sofia University, Tzar Osvoboditel Bld. 15,
1500 Sofia, Bulgaria
2
CEMP, Department of Geography, Lancaster University, Lancaster LA1 4YB,
United Kingdom; v.karloukovski@lancs.ac.uk
3
Geological Institute, Bulgarian Academy of Sciences, Acad. G. Bonchev Str. 24 Bl.,
1113 Sofia, Bulgaria; zlatkam@geology.bas.bg, avhar@geology.bas.bg
4
Manager, SURRC/NERC Argon Isotope Facility, Scottish Universities Research and Reactor Centre,
Scottish Enterprise Technology Park, Rankine Avenue, East Kilbride G75 0QF, Scotland, United Kingdom; m.pringle@surrc.gla.ac.uk
(Manuscript received December 11, 2002; accepted in revised form October 2, 2003)
Abstract: Two multiple stratified volcano-sedimentary Paleogene sections built of breccia-conglomerates and sand-
stones (incl. epiclastic), marls, limestones (incl. reef and bioclastic) and zeolitized pyroclastic rocks (interpreted as
pyroclastic flow deposits mainly) were studied and correlated. The deposits are divided into superposed and interfingered
informal lithostratigraphic units. The natural remanent magnetization of the zeolitized pyroclastics was determined as
thermal remanent magnetization (TRM) or partial TRM, and that of the sediments as detrital remanent magnetization
and post-detrital remanent magnetization. The characteristic remanent directions allowed straightforward construction
of magnetostratigraphic profiles and yielded 12 reliable magnetic polarity zones and sequences of zones in the sections
studied. The correlation of the sections was based on the magnetic polarity zones position, lithological features of the
rocks and three
40
Ar/
39
Ar age determinations of the zeolitized pyroclastics from different levels of the sections. The
consistent (Rupelian 32.28±0.07 to 31.82±0.07) ages plot entirely into the 12r polarity chron on the Geomagnetic
Polarity Time Scale (GPTS) (Cande & Kent 1995). The data presented show that the polystage formation of the pyroclas-
tic flows and the change of the contemporaneous sedimentation from clastic to biogenic took place in a relatively short
time span of about 0.46 Ma.
Key words: Paleogene, Bulgaria, Eastern Rhodopes, correlation, lithology, magnetostratigraphy,
40
Ar/
39
Ar age.
Introduction
During Late Eocene (Priabonian)Oligocene the territory of
South Bulgaria was affected by intensive processes of differ-
ential block movements, contemporaneous terrigenous (conti-
nental and marine) sedimentation and orogenic, collision relat-
ed, magmatism (Dabovski et al. 1989). At the end of the
Priabonian and during the Rupelian the Eastern Rhodopes
were a complex archipelago in a vast shelf. Most of the islands
were of volcanic origin. Apart from the tectonic regime, two
groups of factors controlled their places, dimensions and
shapes: a) polystage and polystyle activity of subaerial volca-
noes composed of acid or/and intermediate volcanics and vol-
caniclastics and b) simultaneous (s.l.) processes of intensive
catastrophic and erosional destruction of the volcanic edifices.
The time and space combination of volcanic processes and
marine sedimentation resulted in an accumulation of various
clastic (incl. epiclastic), biogenic (incl. reef limestones), pyro-
clastic and diverse volcano-sedimentary deposits. In their rela-
tionships a dominating superposition is combined with com-
plex interfingering and gradual lateral transition. Such
lithological variability poses serious problems both from the
litho- and chronostratigraphic points of view, which have not
been solved satisfactorily on the basis of paleontological data
only.
The aim of the investigations reported is to correlate two
Eastern Rhodopes volcano-sedimentary sections using litho-
logical, magnetostratigraphic and radiometric (
40
Ar/
39
Ar)
methods.
Geological setting and lithology of the sections
studied
Two multiple Paleogene sections (A and B) from the Mom-
chilgrad-Arda magmatic region were studied (ϕ = 41.5° N,
λ = 25.5° E). Section A is located to the west of Momchilgrad
(Fig. 1) and consists of six segments situated: to the west-
northwest of Sofiytsi, to the east of Vurhari, to the west and
southwest of Gradinka, to the southwest of Sedlari, to the
southeast of Slunchogled and to the east of Tyutyunche (the
last lies outside the territory presented on Fig. 1). Section B is
located to the southeast of Momchilgrad, consisting of four
segments to the west of Kos, to the southeast of Lale, to the
north of Chayka and to the west-northwest of Karamfil (Fig.
2). The well stratified sedimentary (incl. epiclastic) and bed-
252 MOSKOVSKI et al.
ded pyroclastic rocks cropping out along the sections are di-
vided into six informal lithostratigraphic units:
Formation of breccia-conglomerates and sandstones
It overlays an uneven surface of Pre-Cambrian (?) gneisses
and serpentinized ultramafic rocks (Fig. 1). The formation is
built by polymict breccia-conglomerates and sandstones. The
breccia-conglomerates (gravel- to cobble-size) are unsorted,
with clasts from the basements metamorphics mainly (gneiss-
es, mica-schists, calc-silicate schists, marbles and metamor-
phic quartz). Well rounded clasts of serpentinized ultramafic
rocks, polymict sandstones, polymict conglomerates and silt-
stones are rarely present. The sandstones are thick-bedded and
Fig. 1. Geological setting of the segments of the multiple section A
(to the west of Momchilgrad). Segment Tyutyunche is located out-
side of the map. The map is according to Moskovski et al. (1990).
Fig. 2. Geological setting of the segments of the multiple section B
(to the south-east of Momchilgrad). The map is according to Moskov-
ski et al. (1990).
CORRELATION OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES 253
fine-grained. They are interbedded by thin siltstone layers.
Their terrigenous component (quartz, feldspars and musco-
vite) is cemented by carbonate mineral.
The thickness of the formation varies from 0.4 to 25 m. Re-
cently this formation was referred to the conglomerate-sand-
stone packet of the Oligocene formation of the first acid vol-
canism (Kozhoukharov et al. 1989).
Formation of zeolitized acid pyroclastics, epiclastics and
limestones
It covers the previous formation nearly sharply (Fig. 1). The
formation consists of an irregular alternation of zeolitized acid
pyroclastics (former pyroclastic flows deposits) and packets of
sedimentary rocks (sandstones, siltstones and limestones),
containing epiclastic volcanic components in strongly variable
quantities. The zeolitized pyroclastics are grey to white, grey-
green, light green or grey-pink and are built mainly of former
vitroclasts (Fig. 3). They also contain crystalloclasts of sani-
dine, acid plagioclase, biotite, cognate phenorhyolitic clasts
and foreign clasts of phenoandesites and metamorphics. The
thickness of the zeolitized pyroclastic packets varies from 10
12 to 65 m. The epiclastic sandstones and siltstones contain
clasts of phenorhyolites, metamorphics and zeolitized pyro-
clastics, together with grains of acid plagioclase, hornblende,
biotite, quartz and sanidine (Fig. 4). Rare bivalvian, foramin-
iferal, echinoidean, algal and bryozoan fragments are also
present. The matrix is polymict, with calcite grains in some
places. The limestones are micrite to biomicrite (acc. to the
Folks 1962 classification) or wackestones (acc. to the Embry
& Klovan 1972 classification), sandy, with the epiclastic com-
ponent dominating (Fig. 5).
The formation is about 150 m thick. On the geological map,
scale 1:100,000 (Kozhoukharov et al. 1989) it is included into
the Oligocene formations of first acid volcanism and of
second intermediate volcanism.
Fig. 3. Zeolitized acid pyroclastic rock: completely zeolitized pum-
ices (zp), crystalloclasts of plagioclase (Pl), sanidine (Sn) and
quartz (Qz). N II. (Segment Sofiytsi, section A.)
Fig. 4. Medium to fine epiclastic sandstone. Unrounded and unsorted
grains of plagioclase (Pl), biotite (Bi) and zeolitized pumice frag-
ments (zp) in microsparite groundmass. N II. (Segment Gradinka, sec-
tion A.)
Fig. 5. Sandy fossiliferous microsparite limestone (wackestone).
Unrounded and unsorted fresh clasts of plagioclase (Pl) and slightly
altered glass shards (g). Biofragments of benthic and planktonic for-
aminifers (f). N II. (Segment Sofiytsi, section A.)
Fig. 6. Zeolitized acid pyroclastic flow deposit. The zeolitized
glass shards and pumice fragments contact by clay (Cl) rim; crys-
talloclasts of plagioclase (Pl). N II. (Segment Gradinka, section A.)
254 MOSKOVSKI et al.
Formation of zeolitized acid pyroclastics
In section A (Fig. 1) it overlays the rocks of the previous
formation conformably and with sharp contact. The zeolitized
rocks (Boschinov 1976; Kirov et al. 1976) in that section are
grey-white, thick-bedded to massive. They were formed as a
result of alteration of ash-size pyroclastics mainly (Figs. 6, 7)
and have a total thickness of about 90 m. It has to be noted
that both their foreign volcanic clasts of intermediate compo-
sition and their acid cognate clasts contain early biotite, en-
closed into pyroxene and plagioclase phenocrystals. This pe-
culiarity is typical for the volcanics from the region of the
Borovitsa caldera, located in the westernmost part of the East-
ern Rhodope Paleogene magmatic area (unpublished data of
P. Marchev in Moskovski et al. 1990). A genetic link between
section As former pyroclastics and the Borovitsa caldera ac-
tivity was also suggested by Yanev (1995) and Raynov et al.
(1997).
To the southeast of Momchilgrad (section B Fig. 2) the
formation was deposited above a series of undivided volca-
nics and volcaniclastics, which were not included into the
sampling program. The zeolitized acid pyroclastics in that
section are up to 140 m thick and are mainly composed of
lapilli, to block varieties (Figs. 8, 9). Their primary pumice
clasts and glass shards are completely replaced by zeolite
minerals clinoptilolite, mordenite, analcime (Djourova &
Aleksiev 1984; Djourova & Boyadjiev 1999). The rocks of
this formation are considered to be part of the packet of rhy-
olitic and rhyodacitic tufaceous breccias, tuffs and tuffites of
the Oligocene formation of the second acid volcanism
(Kozhoukharov et al. 1989, 1992).
Sandstone formation
It is known in the Bulgarian literature as Djebel sand-
stones. The formation is presented in section A only (Fig. 1)
where it overlays conformably the formation of the zeolitized
acid pyroclastics. Their lowermost 10 m consist of grey-
greenish epiclastic sandstones and siltstones, followed up-
ward by quartz-feldspar sandstones, interbedded by sandy
clays and siltstones (Bozhinov 1981). As a whole the rocks of
the sandstone formation are loose and because of this only the
lowermost 30 meters of their section were sampled for paleo-
magnetic studies. The lower level of the formation is referred
to the Lower Oligocene on the basis of macrofauna assem-
blage (Sapoundjieva & Yanev 1984).
Fig. 7. Former pyroclastic flow deposit. Densely touched zeolitized
glass fragments (zg), crystalloclasts of plagioclase (Pl) and de-
formed biotite (Bi). N II. (Segment Slunchogled, section A.)
Fig. 8. Zeolitized pyroclastic flow deposit rich in cognate rhyolitic
clasts (Rh) and zeolitized pumice fragments (ZP). The largest pum-
ice is about 2.5 cm long. (Segment Chayka, section B.)
Fig. 9. Microphotograph of the sample of Fig. 8. The pores between
the clasts are filled by zeolite crystals (z). N II.
Fig. 10. Bedded epiclastic sandstones near Karamfil. (Segment
Karamfil, section B.)
CORRELATION OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES 255
Formation of phenoandesite epiclastics (Moskovski et al.
1993; Harkovska et al. 1994)
This formation is presented in section B only. It overlays
the formation of zeolitized pyroclastics (Fig. 2). Grey to yel-
lowish thick bedded to massive sandstones and siltstones are
the main rock types (Fig. 10). The sandstones from the lower
levels of the formation pass upward into fine-grained sand-
stones and siltstones. In the upper levels the sandstones are in-
terbedded with lenses of unsorted conglomerates with well-
rounded pebble- and cobble-size clasts of intermediate
volcanics (phenoandesites?). The presence of sparse small and
undeformed pillows of intermediate to basic composition is
another typical feature of the upper levels of this formation.
Grains of acid plagioclase and clasts of phenorhyolites, rarely
of phenoandesites, pyroxene and amphibole, are also distin-
guished as terrigenous components in the epiclastic sand-
stones (Fig. 11). Fragments of planktonic foraminifers and di-
atoms are also present. In some places the polymict matrix is
replaced by calcite groundmass. The thickness of the forma-
tion is about 50 m. Some years ago the age of its lowermost
levels was dated as Chattian (nannoplankton Zones NP24
NP25, Harkovska et al. 1998). On the geological map (scale
1:100,000) the rocks of the formation were referred to the Oli-
gocene formation of the third intermediate volcanism
(Kozhoukharov et al. 1992).
Formation of marls and bioclastic limestones (Moskovski et
al. 1993)
This is a locally developed formation, which is presented in
the section B only as a sequence overlaying the formation of
phenoandesite epiclastics (Fig. 2). Dark grey sandy to silty
fine-bedded marls are its main rock type. The marls are inter-
bedded by lenses or thin beds of bioclastic limestones, sand-
stones and polymict conglomerates (Figs. 12, 13). This sedi-
mentary sequence is about 6575 m thick. On the recent
geological map (scale 1:100,000) it is included in the forma-
tion of the third intermediate volcanism (Kozhoukharov et
al. 1992). The paleontological age of the sequence is a contro-
versial one: Priabonian according to benthic foraminiferal
assemblage (Harkovska et al. 1992) or Rupelian (unpublished
nannoplankton determination of Kr. Stoykova). The undivid-
ed acid zeolitized pyroclastics, which crop out above this for-
mation were not sampled for paleomagnetic studies, because
they belong to a paleolandslide (Harkovska & Djourova
1994).
Magnetostratigraphy
A total of 434 oriented samples were collected from the 429
meters of section A and the 340 meters of section B, their po-
sitions marked as ticks to the right of the magnetic polarity
profiles on Fig. 16. The average spacing between sampling
levels was 2.2 m (section A) and 2.3 m (section B). The sam-
ples were divided into three main groups: a) zeolitized acid
pyroclastics; b) epiclastic sandstones and siltstones; and c)
marls and bioclastic limestones. Their natural remanent mag-
Fig. 11. Coarse to medium epiclastic sandstone. Unsorted and un-
rounded clasts of plagioclase (Pl), phenorhyolite (Rh) and zeoli-
tized glass (zg). The matrix is polymict, partially replaced by mi-
crosparite. N II. (Segment Karamfil, section B.)
Fig. 12. Alternation of marls and bioclastic limestones in segment
Kos (section B).
Fig. 13. Biomicrite limestone (wackestone to packstone) with bryo-
zoan (br), algal (al) and foraminiferal (f) biofragments and rare pla-
gioclase (Pl) clasts. N II. (Segment Kos, section B.)
256 MOSKOVSKI et al.
netization (NRM) was measured at the School of Environ-
mental Sciences, University of East Anglia (Norwich) with
spinner and cryogenic magnetometers. On the basis of their
relatively high Koenigsberg ratios Q
n
(average Q
n
= 0.95), the
NRM of the first group was most probably a thermorema-
nence (TRM) or, as in the case of the pyroclastics from seg-
ments Karamfil, Lale and Chayka, which contain abundant
foreign clasts a partial TRM. Alternating field (AF) de-
magnetizations of these pyroclastics were completely chaotic,
while the thermal ones revealed linearity up to a certain point
only the point of the emplacement temperature. With aver-
age Q
n
s of 0.43 and 0.24, respectively, the NRM of the epi-
clastic rocks, marls and limestones was detrital (DRM) or
post-detrital (PDRM).
A strong viscous overprint of normal polarity was frequent-
ly revealed in stepwise AF or high-temperature demagnetiza-
tion in all rocks. It was cleaned after 200250 °C or 20
25 mT AF field (Fig. 14a), after which point straight linear
segments on the Zijderveld plots (Zijderveld 1967) deter-
mined unambiguously characteristic remanent magnetization
(ChRM) directions of either normal or reversed polarity. The
samples ChRM directions were predominantly normal or re-
versed. In single cases, transitional polarities were isolated,
and these were excluded from the analysis. The thermal de-
magnetization of the NRMs and of three-component isother-
mal remanent magnetization (Lowrie 1990) revealed two
main unblocking temperatures one, around 300400 °C,
corresponding to medium titanomagnetites (TM30TM40)
(Heller et al. 1979), and a second, magnetite-like temperature
of 550570 °C (Fig. 14b).
Hysteresis measurements in fields up to 1 T on a vibrating
sample magnetometer, modified to additionally measure
back-field curves, identified predominantly multidomain
(MD) and some pseudo-single domain (PSD) carriers of rema-
nence in the zeolitized acid pyroclastics (J
rs
/J
s
= 0.030.2, H
cr
/
H
c
= 2.512) (Day et al. 1977). The sedimentary rocks were
magnetically harder, with mainly PSD and some MD medium
titanomagnetites and magnetites (J
rs
/J
s
=0.040.3, H
cr
/H
c
=28).
The demagnetization characteristics of 125 (29 %) of all py-
roclastic and sediment samples were of quality (linearity,
number of demagnetization points and span of the segment
used to calculate the ChRM) high enough to be comparable to
that of the intermediate and acid volcanic bodies from the area
between sections A and B studied alongside them (Karlou-
kovski 2000). These pyroclastic and sediment samples came
from four segments to the west of Momchilgrad and from one
segment to the east of Momchilgrad. Their site-averaged
ChRM directions are given in Table 1 and are plotted on Fig.
15. The mean positive and the mean negative ChRM directions
(D = 6.5°, I = 54.8°, α
95
= 14.4° and D = 182.7°, I=53.3°,
α
95
= 9.7° correspondingly) pass McFadden & Elhinneys
(1990) reversal test and are essentially antipodal (Karloukovs-
ki 2000). The antipodality is an important indicator of the ade-
quate removal of secondary overprints and of the primary na-
ture of the isolated ChRMs, especially in view of the
closeness of the mean positive ChRM direction to the modern
axial dipole field for the area (D = 0°, I = 60.5°) (Fig. 15). The
mean ChRM direction from all sediments and pyroclastics (D =
184.5°, I = 54.1°, α
95
= 7.0°) is close to the mean direction
Fig. 14. Typical AF (a) and thermal (b) demagnetization characteristics of NRM. Zeolitized acid pyroclastics (sample D261, segment Sed-
lari, section A) were used in (a), and epiclastic sandstones (sample D291, segment Vurhari, section A) in (b).
Segment
N
D (º)
I (º) á
95
(º)
Kos (marls, limestones)
14
0.1
39.7
10.8
Lulichka (marls, limestones)
6
18.1
47.3
13.9
Sedlari (pyroclastics)
13
171.4
50.6
9.5
Vurhari, negative (pyroclastics)
22
178.3
52.2
19.5
Vurhari, positive (pyroclastics)
5
7.1
64.4
14.3
Gradinka, negative (pyroclastics) 44
189.5
54.4
4.5
Gradinka, positive (pyroclastics) 21
4.6
62.5
6.4
Mean (negative)
3
182.7
53.3
9.7
Mean (positive)
4
6.5
54.8
14.4
Mean (all)
7
184.5
54.1
7.0
Table 1: Site mean directions in the pyroclastics and the sediments.
Segment Lulichka is located to the south of Karamfil (Fig. 2), out-
side the map. N number of samples; D, I site mean declination
and inclination; α
95
confidence limit.
CORRELATION OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES 257
(D = 182.9°, I = 51.4°, α
95
= 5.2°, 12 sites) obtained from the
neighbouring lava rocks (Karloukovski 2000). This would
suggest a PDRM rather than a DRM origin of the NRMs in
the sediments due to the lack of inclination shallowing.
The magnetic polarity section between the radiometric data
(32.28 to 31.82 Ma) includes at least 7 polarity events, five of
which (N4 to N6, N8 and N9) are securely attested in multiple
sections. Some of them, like N6, are most certainly polarity
excursions and could not have been registered on the GPTS.
Others are 10 to 20 meters high (N4, N8 in sediments) and
probably 10 to 20 ka in duration. We think that some of them
correspond to the three registered cryptochrons of normal po-
larity in the 12r chron C12r-5, C12r-4, C12r-3 (Cande &
Kent 1995).
The ChRM directions of the sediments and pyroclastics al-
lowed straightforward construction of the magnetostratigraph-
ic profiles, and yielded reliable magnetic polarity zones and
sequences of zones in the two multiple sections studied (the
normal zones being numbered through N1 to N12 from bot-
tom to top) (Fig. 16).
Radiometric data
Three samples of zeolitized acid pyroclastic rocks from the
both sections studied were crushed, sieved and passed
through Franz separator, before single crystals of sanidine and
biotite could be hand-picked. The samples were analysed at
SURRC in East Kilbride, UK using a mass spectrometer and
gas line devoted to low radiogenic argon content rocks. Step-
heating experiments were combined with single-crystal fusion
analyses for best total average age. Analytical procedures
were similar to those used by Singer et al. (1999). The sam-
ples, along with 27.92 Ma sanidine standard Tcr-2a, were
placed in quartz vials and irradiated at the Oregon State Uni-
versity Triga reactor in the cadmium lined in-core irradiation
tube (CLICIT). Prior to the analyses, samples were degassed
in a vacuum for approximately 2 weeks to remove any ab-
sorbed air. To release the radiogenic and potassium-derived
argon, individual samples were heated to progressively higher
temperatures in a double vacuum furnace from around 500 to
1200 °C over 912 steps. Reactive gases were removed using
a two-stage gas clean up, incorporating hot (450 °C) and cold
(room temperature) ZrAl getters. A zeolite cold finger was
used to trap hydrocarbons. Relative amounts of the different
argon isotopes in the cleaned gas fraction were measured with
a MAP-215 spectrometer. The
40
Ar/
39
Ar ratios used to calcu-
late the age for each step were corrected for Ca-derived
39
Ar,
air-derived
40
Ar, the equipment blank and mass spectrometer
mass discrimination. The analytical errors of all these quanti-
ties were included in the standard deviation. Criteria for se-
lecting reliable results have been established by Singer &
Pringle (1996).
All the three ages obtained plot onto the Rupelian time span
of the Gradstein & Oggs (1996) timescale Table 2.
Correlations
The following correlations of the sections studied could be
made using the data reported:
1. The specific alternation of the normal polarity zones from
N7 to N9 (Fig. 16), revealed in the both sections, led to their
secure leveling (Karloukovski 2000) and could be used as a
marker in the magnetostratigraphic correlations:
a) The lowermost levels of the sandstone formation in sec-
tion A (epiclastic rocks in segments Tyutyunche and Slun-
chogled) and the lowermost levels of the formation of phe-
noandesite epiclastics in section B (segment Karamfil) have to
be correlated because they belong to the N8 polarity zone.
These levels are also very well comparable from the lithologi-
cal point of view.
b) The uppermost 40 m of the formation of zeolitized acid
pyroclastics in section A and the whole 140 m of the same
formation in section B have to be correlated not only on the
basis of the magnetostratigraphic data (both intervals belong
Fig. 15. Stereoplot of the site-mean directions in the sediments and
pyroclastics. The star indicates the direction of the present dipole
magnetic field. Open and closed symbols denote the lower and up-
per hemisphere projections, correspondingly.
Sample
Number
Mineral
Number of
crystals
Number of
analyses
Age
(Ma)
I
(Ma)
Section and
segment
Geological age according to
Gradstein & Ogg (1996)
D337
Sanidine
6
15
32.28
±0.07
A, Vurhari
Rupelian
D227
Sanidine
6
10
32.00
±0.06
A, Sedlari
Rupelian
D462
Biotite
6
6
31.82
±0.07
B, Karamfil
Rupelian
I standard deviation of the normal distribution
Table 2:
40
Ar/
39
Ar ages of zeolitized acid pyroclastics.
258 MOSKOVSKI et al.
Fig. 16. Magnetostratigraphy and proposed correlation of the sections studied. The normal polarity zones on the magnetostratigraphic col-
umns are shown in black and are numbered from N1 to N12, the reverse ones in white, the transitional zones are hatched.
CORRELATION OF PALEOGENE SECTIONS IN THE EASTERN RHODOPES 259
Fig. 17. Position of the combined magnetic polarity profile of the sections stud-
ied on the Geomagnetic Polarity Time Scale (Cande & Kent 1995) and on the
Geological Timescale of Gradstein & Ogg (1996).
to the N7 polarity zone), but according to their
40
Ar/
39
Ar ages and lithological features as well.
2. On the Geomagnetic Polarity Time Scale
(GPTS) the dated part of section A (between sam-
ples D337 and D227), which corresponds to a time
span of 0.46 Ma, falls entirely within the lower
part of the 12r polarity chron (Fig. 17). This result
is in agreement with the paleomagnetic data on
previously studied volcanic bodies from a territory
situated between the sections studied (Dola-
pchieva et al. 1986).
3. The
40
Ar/
39
Ar age of biotite (31.8±0.07 Ma
Sp. No. D462, segment Karamfil, section B;
Table 2) is consistent with the
40
Ar/
39
Ar age of
plagioclase (31.39±0.50 Ma) from similar pyro-
clastics near Karamfil (Singer & Marchev 2000).
The cited authors interpreted these pyroclastics as
outflow tuffs of the Borovitsa caldera. The same
origin could be suggested for all the zeolitized py-
roclastic flows deposits in section A, having in
mind their age and the common petrographic fea-
tures of their foreign and cognate clasts on one
hand and the intermediate and acid volcanics from
the Borovitsa volcanic region on the other.
4. The differences in the dimensions of the
former pyroclasts and cognate clasts in the zeoli-
tized acid pyroclastics in section A as compared
with those in section B could be explained with a
suggestion that the respective pyroclastic flow de-
posits were related to different explosive events in
the Borovitsa caldera realm.
Conclusions
1. The study shows that the combination of
lithostratigraphic, magnetostratigraphic and high-
resolution radiometric (
40
Ar/
39
Ar) data can be a
very useful tool for correlation of complex strati-
fied volcano-sedimentary sections and for discrim-
ination of the respective volcanic events and pro-
cesses:
a) Two lithostratigraphic units of different li-
thology (sandstone formation section A and
formation of epiclastics section B) were com-
pletely correlated.
b) Two lithostratigraphic units of completely
comparable lithology (formation of zeolitized acid
pyroclastics) were partially only correlated.
2. The polystage formation of pyroclastic flows,
their zeolitization into the Rupelian marine basin
and the contemporaneous sedimentation took
place in a relatively short time span of approxi-
mately 0.46 Ma, as is shown by the
40
Ar/
39
Ar dat-
ing of different levels of section A.
260 MOSKOVSKI et al.
3. The accumulation of the rocks in the area of section A is
a result of sedimentation and contemporaneous acid explosive
volcanic activity in the Borovitsa caldera realm. Volcanic
products related to centre(s) of intermediate composition
(mentioned by many previous authors) were not recorded in
the territory in question.
Acknowledgments: The paleomagnetic and
40
Ar/
39
Ar studies
were financed by Lord Zuckermanns studentship at the Uni-
versity of East Anglia, Norwich. The geological maps (Fig. 1
and Fig. 2) are from the report of the Project 16604-K (Sub-
project 2), Sofia University. The authors would like to express
their gratitude to the reviewers Dr. N. Yordanova, Dr. A. Di
Stefano and Dr. I. Túnyi for the critical and helpful comments.
Y. Zagorcheva drew figures 1 and 2.
This paper was presented at the XVIIth Congress of Car-
pathian-Balkan Geological Association held in Bratislava,
SR, in September 2002.
References
Boschinov K. 1976: Über die Ausdehnung und über die mineralche-
mische Charakteristik einer Schicht Öligozänisch-vulkanischer
Tuffe von den Südöstlichen Rhodopen. Ann. de lUniv. de So-
fia Kl. Ohridski, Ser. 1 Geol. 68, 1, 279288 (in Bulgari-
an, German abstract).
Bozhinov K. 1981: Composition of the Djebel sandstones. Rev.
Bulg. Geol. Soc. 42, 1, 130137 (in Bulgarian).
Cande S. & Kent D. 1995: Revised calibration of the geomagnetic
polarity time scale for the Late Cretaceous and Cenozoic. J.
Geophys. Res. 100, 60936095.
Dabovski C., Harkovska A., Kamenov B., Mavroudchiev B., Stani-
sheva-Vassileva G. & Yanev Y. 1989: Geodynamic model of
the Alpine magmatism in Bulgaria. XIV Congress CBGA. Re-
ports, Sofia, 4466.
Day R., Fuller M. & Schmidt V. 1977: Hysteresis properties of tita-
nomagnetites: grain size and compositional dependence. Phys.
Earth Planet. Inter. 13, 260267.
Djourova E. & Aleksiev B. 1984: Zeolitic rocks deposits of eco-
nomic value in the North-Eastern Rhodopes. Ann. de lUniv. de
Sofia Kl. Ohridski, L. 1 Geol. 78, 256262 (in Bulgarian,
English abstract).
Djourova E. & Boyadjiev S. 1999: The zeolitic rocks in the vicinity
of Karamfil village, Haskovo district mineralopetrographic
and geochemical characteristics. Geology and Mineral Re-
sources 7, 1823 (in Bulgarian, English abstract).
Dolapchieva M., Nozharov P. & Petkov N. 1986: Paleomagnetic
characteristics of the young volcanics of the Eastern-Rhodope-
an depression. Bulg. Geophysical J. 12, 2, 114126 (in Bulgar-
ian, English abstract).
Embry E. & Klovan J. 1972: Absolute water depth limits of late De-
vonian paleoecological zones. Geol. Rdsch. 61, 672686.
Folk R. 1962: Spectral subdivision of limestone types. In: Ham
W.E. (Ed.): Classification of carbonate rocks. Amer. Assoc.
Petrol. Geol. Mem. 1, 6284.
Gradstein F. & Ogg J. 1996: Geological timescale. Episodes 19, 12.
Harkovska A. & Djourova E. 1994: Petrologic features and slump
redeposition of Paleogene rhyolitic tuffs from Eastern
Rhodopes (Bulgaria). Int. Volcanol. Congr. IAVCEI 1994, An-
kara, Abstracts, 169.
Harkovska A., Moskovski S. & Jouranov S. 1992: Upper Eocene
(Priabonian) rhyolite tuffs in the region of Dambaluk Mountain
(East Rhodopes, Bulgaria). Geol. Balcanica 22, 1, 74.
Harkovska A., Moskovski S. & Nedyalkov R. 1994: Paleogene vol-
canomictic sediments (epiclastics) in the south-eastern parts of
the Momchilgrad-Ardino volcanic region (SE Bulgaria). C. R.
Acad. Bulg. Sci. 47, 10, 5356.
Harkovska A., Pe-Piper G., Stoykova K., Nedyalkov R. & Mosk-
ovski S. 1998: Late Oligocene (Chattian) epiclastics and lava
flows of intermediate composition from the Eastern Rhodopes
(Bulgaria). C. R. Acad. Bulg. Sci. 51, 78, 5358.
Heller F., Markert H. & Schmidtbauer E. 1979: Partial self-reversal
of NRM of an historical lava flow of Mt. Etna (Sicily). J. Geo-
phys. 45, 235257.
Karloukovski V. 2000: Magnetostratigraphy and palaeomagnetism
of the area around the Momchilgrad Paleogene depression, the
East Rhodope massif. PhD Thesis, University of East Anglia,
Norwich, 1343.
Kirov G., Pechigargov V. & Chelebiev E. 1976: Effect of the grain
size of volcanic glass on the mineral composition of its alter-
ation products. Geochemistry, Mineralogy and Petrology, 4 ,
8390 (in Bulgarian, English abstract).
Kozhoukharov D., Boyanov I., Goranov A., Yanev Y., Shilyafova J.
& Rousseva M. 1989: Geological map of Bulgaria, Scale
1:100,000, Sheet Kurdjali, Sofia.
Kozhoukharov D., Boyanov I., Goranov A. & Kozhoukharova E.
1992: Geological map of Bulgaria, Scale 1:100,000, Sheet
Kroumovgrad and Sape, Sofia.
Lowrie W. 1990: Identification of ferromagnetic minerals in a rock
by coercivity and unblocking temperature properties. Geophys.
Res. Lett. 17, 159162.
McFadden P. & McElhinny M. 1990: Classification of the reversal
test in paleomagnetism. Geophys. J. Int. 103, 725729.
Moskovski S., Harkovska A. & Marchev P. 1993: New data on the
stratigraphy of the Paleogene in the Dambaluk volcanic massif.
C. R. Acad. Bulg. Sci. 46, 6, 6164.
Moskovski S., Harkovska A., Marchev P., Jouranov S. & Lilov P.
1990: Stratigraphic, petrologic and structural-volcanologic
studies in the Dambaluk volcanic massif. Project 16604-K,
Subproject 2. Fac. of Geol. and Geogr., Sofia University, Sofia
(in Bulgarian, unpublished).
Raynov N., Popov N., Yanev Y., Petrova P., Popova T., Hristova V.,
Atanasova R. & Zankarska R. 1997: Geological, mineralogical
and technological characteristics of zeolitized (clinop-
tilolitized) tuff deposits in the Eastern Rhodopes, Bulgaria. In:
Kirov G., Filizova L. & Petrov O. (Eds.): Natural zeolites. So-
fia95. PENSOFT, SofiaMoscow, 263275.
Sapoundjieva V. & Yanev Y. 1984: Fauna from the sandstones near
Djebel, East Rhodopes. Rev. Bulg. Geol. Soc. 45, 308317 (in
Bulgarian, English abstract).
Singer B. & Marchev P. 2000: Temporal evolution of arc magma-
tism and hydrothermal activity, including epithermal gold
veins, Borovitsa caldera, Southern Bulgaria. Econ. Geol. 95,
11551164.
Singer B., Hoffman K., Chauvin A., Coe R. & Pringle M. 1999: Dat-
ing transitionally magnetised lavas of the late Matuyama
Chron: toward a new
40
Ar/
39
Ar timescale of reversals and
events. J. Geophys. Res. 104, 679693.
Singer B. & Pringle M. 1996: Age and duration of the Matuyama
Brunhes geomagnetic polarity reversal event from
40
Ar/
39
Ar
incremental heating analyses of lavas. Earth Planet. Sci. Lett.
139, 4761.
Yanev Y. 1995: General characteristics of the Late Paleogene colli-
sion volcanism in the Rhodopes. In: Aleksiev B. (Ed.): Sofia
Zeolite Meeting 95, International Symposium and Exhibition
on Natural Zeolites, June 1825, 1995, Sofia, Bulgaria. Guide
to the Post-Meeting Field Trip, 319.
Zijderveld J. 1967: A.C. demagnetization of rocks: Analysis of re-
sults. In: Collinson D., Creer K. & Runcorn S. (Eds.): Methods
in palaeomagnetism. Elsevier, Amsterdam, 254286.