Geochronology of Eruptions and Parental Magma Sources of Elbrus Volcano, the Greater Caucasus: K-Ar and Sr-Nd-Pb Isotope Data

Abstract

Complex geochronological and isotope-geochemical studies showed that the Late Quaternary Elbrus volcano (Greater Caucasus) experienced long (approximately 200 ka) discrete evolution, with protracted periods of igneous quiescence (approximately 50 ka) between large-scale eruptions. The volcanic activity of Elbrus is subdivided into three phases: MiddleNeopleistocene (225–170 ka), Late Neopleistocene (110–70 ka), and Late Neopleistocene-Holocene (less than 35 ka).
Petrogeochemical and isotope (Sr-Nd-Pb) signatures of Elbrus lavas point to their mantle-crustal origin. It was shown that hybrid parental magmas of the volcano were formed due to mixing and/or contamination of deep-seated mantle melts by Paleozoic upper crustal material of the Greater Caucasus. Mantle reservoir that participated in the genesis of Elbrus lavas as well as most other Neogene-Quaternary magmatic rocks of Caucasus was represented by the lower mantle “Caucasus” source. Primary melts generated by this source in composition corresponded to K-Na subalkali basalts with the following isotopic characteristics: 87Sr/86Sr = 0.7041 ± 0.0001, ƒNd = +4.1 ± 0.2, 147Sm/144Nd = 0.105–0.114, 206Pb/204Pb = 18.72, 207Pb/204Pb = 15.62, and 208Pb/204Pb = 38.78. The temporal evolution of isotope characteristics for lavas of Elbrus volcano is well described by a Sr-Nd mixing hyperbole between “Caucasus” source and estimated average composition of the Paleozoic upper crust of the Greater Caucasus. It was shown that, with time, the proportions of mantle material in the parental magmas of Elbrus gently increased: from ∼60% at the Middle-Neopleistocene phase of activity to ∼80% at the Late Neopleistocene-Holocene phase, which indicates an increase of the activity of deep-seated source at decreasing input of crustal melts or contamination with time. Unraveled evolution of the volcano with discrete eruption events, lacking signs of cessation of the Late Neopleistocene-Holocene phase, increasing contribution of deep-seated mantle source in the genesis of Elbrus lavas with time as deduced from isotope-geochemical data, as well as numerous geophysical and geological evidence indicate that Elbrus is a potentially active volcano and its eruptions may be resumed. Possible scenarios were proposed for evolution of the volcano, if its eruptive activity were to continue.

Full text

ISSN 00167029, Geochemistry International, 2010, Vol. 48, No. 1, pp. 41–67. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © V.A. Lebedev, I.V. Chernyshev, A.V. Chugaev, Yu.V. Gol’tsman, E.D. Bairova, 2010, published in Geokhimiya, 2010, Vol. 48, No. 1, pp. 45–73.
41
INTRODUCTION
The Greater Caucasus is the only area in the Euro
pean part of Russia that exhibited Quaternary volcanic
activity expressed in the formation of large volcanic
centers: Elbrus and Kazbek. The Elbrus center is situ
ated completely in Russia, on the border region
between two densely populated republics of the North
Caucasus: KabardinoBalkaria and Karachai
Cherkess. Mount Elbrus (5642 m), the highest peak of
Europe and Russia, is located in the southwestern part
of the Elbrus center. This largest dormant volcano in
the European part of Russia is potentially hazardous in
terms of the resumption of catastrophic eruptions [1,
and others]. The emergence of Elbrus and drastic
thawing of its glaciers may lead to heavy ecological
consequences for natural ecosystems, population, and
economics of the North Caucasian region. Therefore,
the problem of possible activation of this volcano has
attracted a considerable attention of Russian geolo
gists, geophysics, and geochemists in the recent
decade.
The reliable prediction of the possible resumption
of volcanic eruptions in the Elbrus region requires
detailed deciphering of the volcanic evolution of this
region, in particular, determination of duration of vol
canic activity, its main phases or periods of maximal
activity, character of volcanism (discrete or continu
ous), dates of last eruptions, and some other temporal
characteristics. At present, this task can be successfully
solved using an isotope–geochronological study of
volcanic products in combination with results of geo
morphological, petrological, geophysical, and other
data. As for any volcano, petrochemical and isotope
geochemical studies play an important role in compre
hensive study of Elbrus. They provide insight into
sources of parental magmas and their petrogenesis,
evolution of parental melts, which allows prediction of
the possible scenario of future eruptions.
Geochronology of Eruptions and Parental Magma Sources of Elbrus
Volcano, the Greater Caucasus: K–Ar and Sr–Nd–Pb Isotope Data
V. A. Lebedev, I. V. Chernyshev, A. V. Chugaev, Yu. V. Gol’tsman, and E. D. Bairova
Institute of Geology of Ore Deposits, Petrography, Mineralogy, and Geochemistry, Russian Academy of Sciences, Staromonetnyi
per. 35, Moscow, 119017 Russia;
email: leb@igem.ru
Received April 23, 2008
Abstract
—Complex geochronological and isotopegeochemical studies showed that the Late Quaternary
Elbrus volcano (Greater Caucasus) experienced long (approximately 200 ka) discrete evolution, with pro
tracted periods of igneous quiescence (approximately 50 ka) between largescale eruptions. The volcanic
activity of Elbrus is subdivided into three phases: MiddleNeopleistocene (225–170 ka), Late Neopleistocene
(110–70 ka), and Late Neopleistocene–Holocene (less than 35 ka).
Petrogeochemical and isotope (Sr–Nd–Pb) signatures of Elbrus lavas point to their mantle–crustal origin.
It was shown that hybrid parental magmas of the volcano were formed due to mixing and/or contamination
of deepseated mantle melts by Paleozoic upper crustal material of the Greater Caucasus.
Mantle reservoir that participated in the genesis of Elbrus lavas as well as most other Neogene–Quaternary
magmatic rocks of Caucasus was represented by the lower mantle “Caucasus” source. Primary melts gener
ated by this source in composition corresponded to K–Na subalkali basalts with the following isotopic char
acteristics:
87
Sr/
86
Sr = 0.7041
±
0.0001, Є
Nd
= +4.1
±
0.2,
147
Sm/
144
Nd = 0.105–0.114,
206
Pb/
204
Pb = 18.72,
207
Pb/
204
Pb
= 15.62, and
208
Pb/
204
Pb
= 38.78. The temporal evolution of isotope characteristics for lavas of
Elbrus volcano is well described by a Sr–Nd mixing hyperbole between “Caucasus” source and estimated
average composition of the Paleozoic upper crust of the Greater Caucasus. It was shown that, with time, the
proportions of mantle material in the parental magmas of Elbrus gently increased: from ~60% at the Middle
Neopleistocene phase of activity to ~80% at the Late Neopleistocene–Holocene phase, which indicates an
increase of the activity of deepseated source at decreasing input of crustal melts or contamination with time.
Unraveled evolution of the volcano with discrete eruption events, lacking signs of cessation of the Late Neo
pleistocene–Holocene phase, increasing contribution of deepseated mantle source in the genesis of Elbrus
lavas with time as deduced from isotope–geochemical data, as well as numerous geophysical and geological
evidence indicate that Elbrus is a potentially active volcano and its eruptions may be resumed. Possible sce
narios were proposed for evolution of the volcano, if its eruptive activity were to continue.
DOI:
10.1134/S0016702910010039

42
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
In this paper, the main phases of Elbrus activity were
distinguished on the basis of generalization of isotope–
geochronological data that were obtained by us for Elbrus
lavas at the Laboratory of Isotope Geochemistry and
Geochronology of the Institute of Ore Geology, Petrog
raphy, Mineralogy, and Geochemistry of the Russian
Academy of Sciences (IGEM RAS). The origin of
parental melts of the volcanics is discussed in light of the
first complex isotope–geochemical study of the Elbrus
magmatic products.
BRIEF INFORMATION ON THE GEOLOGICAL
STRUCTURE OF ELBRUS VOLCANO
Elbrus volcano is located in the central part of the
Greater Caucasus mountainous system, on the north
ern slope of the Main Caucasus Range, in the water
shed of the Black and Caspian sea basins. It was
formed at the final, Quaternary stage of the Late Cen
ozoic magmatism in the Caucasian segment of the
Alpine fold belt, which developed in this region during
the last 10 Mg. The largescale magmatic activity in
this region was provided by mantle “hot spot” activity
that existed since the Late Miocene to the present
against a background of ongoing convergence between
the Arabian and Eurasian lithospheric plates and sev
eral microplates and terranes [2, and others]. Geody
namic and geotectonic settings of Neogene–Quater
nary volcanism within the Caucasus led to the wide
chemical variations in composition of volcanic prod
ucts, diversity in types of volcanic activity, and involve
ment of mantle and crustal sources in magma forma
tion.
Three neovolcanic areas (Kazbek, Central Geor
gian, and Elbrus) are traditionally distinguished within
the volcanic province of the Greater Caucasus. Qua
ternary volcanism of the Elbrus area was manifested by
the formation of the Elbrus volcanic center in its
southwestern part. The Elbrus volcanic center is local
ized on the northern slope of the Main Caucasus
Range, in the watershed of the Malka, Baksan, and
Kuban rivers. lt includes the Elbrus volcano and a
number of small volcanic edifices along its periphery
(PaleoElbrus, Chukhchur, ChomartKol, Syltran,
TashTebe, Tashlysyrt, and Tyzyl). The first magmatic
phase of activity of the Elbrus center was related to the
ejection of basic lavas in the Tyzyl River valley approx
imately 900 ka [3]. Subsequent 800–700 ka were char
acterized by mainly explosive activity of PaleoElbrus,
ChomartKol, Chuchkhur, and possibly another
apparatus with the production of acid pyroclastic
rocks and, more rarely, trachyandesite lavas [1]. Late
Quaternary magmatic activity of the Elbrus center was
marked by eruptions of Elbrus volcano, as well as
TashTebe satellite volcano in the upper reaches of the
Khudes River, and supposedly, edifices on the Syltran
range.
The twocone Elbrus stratovolcano ejected moder
ately acid lavas. It is confined to the junction zone of
sublatitudinal Pshekish–Tyrnyauz suture with the
transverse Elbrus fault (Fig. 1) and its diameter of base
is ~18 km in average. The basement of the volcano is
made up by metamorphic rocks of the Maker Group
and Paleozoic granitoids, which crop out to altitudes
of 3000–3900 m. Thus, the relative height of the vol
canic edifice accounts for 2000—2500 m. The diame
ter of Elbrus at a height of 5300 m is 1.2–1.5 km; from
this height, it emerges at a height of 300–350 m as two
independent Western and Eastern cones with a com
mon basement (altitudes of 5642 and 5621 m, respec
tively). The western slopes of the volcano are bounded
by vertical scarps (Kyukyurtlyu and UlluKam walls)
with remains of volcanic edifices of its precursor,
PaleoElbrus [1, 4].
The Eastern summit of Elbrus is characterized by
excellent preservation of initial volcanic morphologies
and represents a welltruncated cone with well
expressed crater up to 80 m deep. The initial morphol
ogy of the Western summit is preserved worse than that
of the eastern cone. Its crater funnel is approximately
500 m wide and 250–300 m deep. The western and
southwestern parts of the western cone are bounded by
almost vertical scarp up to 200 m high. Elbrus is the
largest center of modern glaciation in the Caucasus.
The entire upper part of the volcanic edifice starting
from altitudes of 3500–4000 m is covered by glaciers
and firn fields. For this reason, only lava flows ejecting
from the volcano on its slopes in all directions are
available for geological study, while its peak is less
accessible.
G.V. Abikh was the first to initiate systematic stud
ies of young volcanic rocks of the Elbrus center [7].
His investigations were continued by such known
geologists as V.V. Dubyanskii, A.P. Gerasimov,
S.P. Solov’ev, K.N. Paffengolts, E.E. Milanovskii,
N.V. Koronovskii, and others. The works of two latter
researchers [5, 8] summed up the first “early” stage in
studying Elbrus. They generalized available geological
data and attempted to divide young volcanics, to dis
tinguish magmatic phases, and to trace the volcanic
evolution of the Elbrus center using geomorphological
and stratigraphic data.
However, the preparation of detailed stratigraphic
schemes and subdivision of volcanogenic sequences at
this stage of research were limited by the absence of
geochronological data due to insufficient development
of isotope methods for reliable dating the Late Quater
nary rocks. The K–Ar datings on some rocks of Elbrus
were obtained for the first time as early as over 30 years
ago and later were generalized in the work of Borsuk
[9]. However, these few data, due to large errors of iso
tope ages, can be considered only as approximate
determinations.
The second “modern” stage of studies of Elbrus
volcano began in the late 1990s. Based on the first reli
able K–Ar and Rb–Sr datings obtained in the Labora
tory of Isotope Geochemistry and Geochronology of
IGEM RAS, it was concluded that Elbrus was formed

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 43
Fig. 1.
Geological map of Elbrus volcano. Complied by V.A. Lebedev using data of field observations and materials of
V.M. Gazeev [4], N.V. Koronovskii [5], and largescale geological maps of the region [6]. (
1
) glaciers and firn fields; (
2–4
) vol
canics of Elbrus: (
2
) Late Neopleistocene–Holocene (<30 ka), (
3
) Late Neopleistocene (110–70 ka), (
4
) Middle Neopleistocene
(220–170 ka); (
5
) pyroclastic rocks and lavas of PaleoElbrus, Chukhchur, Chomartkol, as well as tuffs and ignimbrites from Mt.
Tuzluk and Irikchat glacier; (
6
) felsitic dikes; and (
7
) isotope sampling localities.
TashTebe
Khudes R.
Ingushli R.
Tuzluk
Malka R.
ShauKol R.
Chuchkhur
ChomartKol R.
KyzylKol R.
Chomartkol
Ml43
Ml36
Ml37 Ml41
Bt17
Bt6
PaleoElbrus
Western Elbrus
Eastern Elbrus
Malka R.
KonushKol R.
E8
43
°
20
´
42
°30´
Irik R.
Terskol R.
Baksan R.
Baidaevo
Cheget
Terskol
Azau
Kuban R
.
UlluEzen R.
Georgia
E7
393
Az22
Az23
Аз25
E4
Az26
E5
E6 Az21
Az28
Az
29
E2
Az31
E3
1993Eа
Az30
Az
19
0246 km
1234 567
Э1

44
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
no earlier than 250 ka ago; earlier, its area was occu
pied by its small precursor, PaleoElbrus, that oper
ated at the Eopleistocene–Neopleistocene boundary
(approximately 800–700 ka) in the modern sources of
the Kuban and Kyukyurtlyu rivers [1, 10, 11]. Elbrus
volcano proper appeared a few hundreds thousands years
after the cessation of the PaleoElbrus activity, and
ejected mainly dacitic lava, which flew mainly in the
northern and southern directions, and, to a lesser extent,
along the western and eastern slopes. We outlined the
time ranges of main volcanic phases and came to a con
clusion of the possible activity of Elbrus up to Holocene
time. Based on petrochemical and isotopegeochemical
data on lavas of Elbrus volcano, we first evaluated the
petrogenesis of Elbrus magmas [2].
Concurrently with our studies of Elbrus volcano,
the group of geologists from the Laboratory of Petrog
raphy of IGEM RAS under the direction of
O.A. Bogatikov, in cooperation with scientists from
some other institutes of the Russian Academy of Sci
ences, carried out the complex geologicalpetrological
study of Elbrus lavas, their stratigraphy, as well as geo
ecological and geophysical studies in the Elbrus region
[12–16, and others]. These works resulted in a deci
phering of the collapsed caldera named as the Elbrus
Caldera from satellite images, the obtaining of several
radiocarbon datings on coals from soils buried on vol
cano slopes, an attempt to unravel the chronology of
Elbrus eruptions using EPR dating, and a compilation
of a new geological map of the volcano.
As is known, the presence of the caldera stage in the
evolution of volcano is an important factor, which,
first, indicates the longterm magmatic activity of the
volcano, as well as considerable volumes of erupted
material at early “precaldera” eruptions. In this rela
tion, the problem whether the caldera stage was
present or absent is of great significance in deciphering
the eruptive chronology of this volcano and prediction
of possible scenarios of its further evolution. Note that
now the “Elbrus caldera” is admitted only by
researchers that published data on its possible exist
ence [13]. In addition, the contours of inferred caldera
structure in the maps published in 1998 [13] and sub
sequent works of researchers from Laboratory of
Petrography of IGEM RAS [for example, 4] strongly
differ; moreover, the gravity anomaly beneath the
Elbrus area [17] is interpreted by them both as the
result of basement subsidence and simultaneously, as
indicating the presence of a subsurface magmatic
chamber beneath the volcano (!). It should be empha
sized that our studies do not support the presence of
caldera in the Elbrus volcano region; the inferences of
the absence of the caldera stage in the evolution of this
volcano were recently made by Koronovskii and Dem
ina [18].
An attempt made by collaborators of Laboratory of
Petrography of IGEM RAS and Mineralogical
Department of Moscow State University to decipher
the evolution of Elbrus volcano from EPR data [14–
16], as alternative to our isotope data, were unsuccess
ful. In the last work, one of the authors of the EPR dat
ing method of magmatic rocks [19], using quartz
bearing volcanics of Elbrus as an example, practically
demonstrated the unsuitability of this method for dat
ing young volcanics and repudiated previous conclu
sions on the age and sequence of lava formation
deduced from EPR data [15].
Below, we consider the results of isotope–geochro
nological and isotopegeochemical studies of Elbrus
lavas, which were used to decipher its evolution and to
determine the possible sources that participated in the
genesis of Elbrus magmas. The petrogenetic models
were proposed to explain temporal evolution of the
melts and several possible scenarios are considered for
the further evolution of magmatic activity in the
Elbrus volcano vicinity.
METHODS
Table 1 lists the main characteristics and localities
of samples used in this work for isotopegeochrono
logical studies. As is seen from this table, the collection
of lavas for our study was mainly taken from the south
ern slope of Elbrus volcano. It is related to the fact
that, unlike other slopes (western and northern) of the
volcanic edifice along which lava also ran down, the
deeply incised canyons of the tributaries of the Baksan
River (Azau, Garabashi, and Terskol) on the southern
slope (Azau valley) recover complete stratigraphic sec
tions of the volcanic products from the earliest to the
final phases of its evolution.
The isotopegeochronological study of the rocks of
Elbrus volcano was conducted using special modifica
tion of the K–Ar method designed in IGEM RAS for
dating young magmatic rocks [10]. Its methodology
and main analytical characteristics are reported in our
work [20].
The content of radiogenic Ar was determined on a
highly sensitive lowblank mass spectrometric com
plex developed on the basis of MI1201 IG (
SELMI
,
Sumy, Ukraine) at the IGEM RAS using the isotope
dilution method with
38
Ar
monoisotope as tracer. The
potassium content was measured by the flame spectro
photometric method on an FPA01 photometer
(ELAMCenter, Russia). In order to avoid a possible
distortion of K–Ar dates due to the presence of
excess
40
Ar
in phenocrysts of volcanics, we analyzed
only groundmass. Total errors in the determination
of K–Ar age (
±
2
σ
) account for from 5 to 100 rel. %
and depend on the content of radiogenic argon in
samples and
40
Ar
rad
/
40
Ar
tot
in the analyzed material.
Their absolute values for certain samples are given in
Table 2 together with age dates.
The isotope composition and contents of Rb, Sr,
Sm, and Nd were measured during isotopegeochro
nological study of volcanics from the Elbrus area,
some other neovolcanic areas of the Greater Cauca
sus, as well as rocks from the Paleozoic basement by

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 45
Table 1. Brief characteristics of the studied lava samples from Elbrus volcano, several other neovolcanic centers and areas of the
Greater Caucasus, as well as rocks of the Paleozoic basement of the Elbrus volcanic center
Sample
no. Rock Geological object Sampling locality
Lavas from the western slope of Elbrus (upper reaches of the BiitikTebe River)
Bt
6
AmfñPxoPxBiPl
dacite Upper lava horizon Right bank of the Biitik–Tebe River, 1.5 km
below the edge of the Biitiktebe glacier
Bt
17
AmfñPxoPxBiPl
dacite Lower lava horizon Area of the Kol’tsevoi pass at the Kyukyurtlyu
and BiitikTebe rivers watershed
Lavas from the northern slope of Elbrus volcano (upper reaches of the Malka River)
Ml
36
AmfcPxoPxBiPl
dacite “Sultan waterfall” lava flow Right side of the KyzykKol River, above the
Sultan waterfall, near the Zhelasu mineral
spring
MI37
QAmfcPxoPxBiPl
dacite “Irakhiktyuz locality” lava flow Upper reaches of the Birdzhalysu River,
Birdzhal locality
Ml
41
AmfBiñPxoPxPl
dacite Lavas of the upper reaches of the
Birdzhalysu River Upper reaches of the third tributary of the
Birdzhalysu River
Ml
43
QAmfBicPxoPxPl
dacite “Sultan waterfall” lava flow Right bank of the KyzylKol River, above Sul
tan waterfall, near the Zhelasu mineral spring,
lacet of the Kislovodsk road
Lavas from the southern slope of Elbrus volcano (Azau valley)
393
E
*
QcPxAmfoPxBiPl
dacite “Ice base” lava flow Ice Base area
1993
E
a*
QcPxAmfBiPl
dacite Lavas of the fourth horizon of the
Azau valley Area of the “Terskol” observatory
Az
19
BiAmfcPxoPxPl
dacite Lavas of third horizon of the Azau
valley Confluence of the Pravaya and Levaya Gara
bashi rivers, right side of the Garabashi valley
Az
21
BicPxoPxPl
dacite Lavas of fifth horizon of the Azau
valley Valley of the Malaya Azau River, right side,
below the “Staryi Krugozor” cable way station
Az
22
AmfcPxBioPxPl
dacite Lavas of the fifth horizon of the
Azau valley, lava flow of the Ma
laya Azau River
Area of the Garabashi cable way station
Az
23
BiAmfcPxoPxPl
dacite
''
Area of the “Mir” cable way station
Az
25
BiAmfoPxPl
dacite Lavas of the first horizon of the
Azau valley Malaya Azau River valley, above “Staryi
Krugozor” cable way station, “Slonov’i zady”
rocks
Az
26
''
Lavas of fifths horizon of the
Azau valley Cliffs above the edge of the Bol’shoi Azau gla
cier
Az
28
QcPxAmfoPxPl
dacite Lavas of the fifth horizon of the
Azau valley, Malaya Azau River
lava flow
Lava “tongue” in the Bol’shaya Azau valley
(above waterfall)
Az
29
AmfcPxoPxPl
dacite
''
Lava “tongue” in the Bol’shaya Azau River
valley (below waterfall)
Az
30
QAmfcPxoPxPl
dacite Lavas of the fourth horizon of the
Azau valley Terskol range, 1.5 km from the “Terskol” ob
servatory toward the Ice Base
Az
31
AmfcPxoPxPl
dacite Lavas of the second horizon of
the Azau valley Volch’i Vorota locality, cliffs with columnar
jointing
E1
BiAmfoPxPl
dacite Lava remain of the first horizon of
the Azau valley Left side of the Terskol River valley, above the
settlement of Terskol
E2
BiAmfcPxoPxPl
dacite Lavas of the third horizon of the
Azau valley Confluence of the Pravaya and Levaya Gara
bashi rivers, right side of the Garabashi valley

46
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
Table 1.
(Contd.)
Sample
no. Rock Geological object Sampling locality
E3
BiAmfoPxPl
dacite Lavas of the first horizon of the
Azau valley Terskol river valley, right side of the Terskol
River, 1 km above the Terskol settlement
E4
BicPxoPxPl
dacite Lavas of the fifth horizon of the
Azau valley Watershed of the Bol’shoi and Malyi Azau
glaciers
E5
'' ''
Staryi Krugozor locality
E6
BiAmfoPxPl
dacite Lavas of the first horizon of the
Azau valley
''
E7
QcPxAmfoPxBiPl
dacite Lavas of the fifth horizon of the
Azau valley (Irik lava flow) Lavas of the watershed of the Irik and Irikchat
glaciers
E8
''
Lavas of the western summit of
Elbrus volcano Western summit of Elbrus volcano
Rocks of some neovolcanic centers and areas of the Greater Caucasus
ZG
4
OPxcPx PlOl
subalkali basalt Central Georgian area, lavas of
the Chiatura volcanic center St. Perevisa, Central part
ZG
5
'' ''
St. Perevisa, western part of the Chiatura de
posit, Perevisa adit
ZG
6
'' ''
EL1
Biotite granite porphyry Eldjurtu Massif Baksan River valley, left side, above Tyrnyauz
MW
5
Mediumgrained biotite granite
porphyry Razvalka Massif, CMW Southeastern slope of Mt. Razvalka
MW
8
Finegrained biotite granite
porphyry Ostraya Massif, CMW Eastern slope of Mt. Ostraya
MW
9
Subvolcanic rhyolite Sheludivaya Massif, CMW Eastern slope of Mt. Sheludivaya
Rocks of the Paleozoic basement of the Elbrus volcanic center
CH21**
Biotite–muscovite–andalusite
schist with garnet Maker Group Mt. Cheget
AZ3**
Mediumgrained biotite granite Paleozoic granitoids of the Main
Caucasus Range Right side of the Bol’shoi Azau glacier
AZ4**
Finegrained leucogranite
'' ''
AZ5**
Biotite–muscovite schist Maker Group Northern slope of Mt. Cheget
SH28**
Mediumgrained biotite granite Paleozoic granitoids of the Main
Caucasus Range Shkhelda ravine
AR403**
Garnet–cordierite gneiss with
sillimanite Metamorphic rocks of the Main
Caucasus Range Adyrsu River Canyon, Lokomotiv peak
72**
Amphibolized eclogite Eclogites of the Chegem Canyon Upper reaches of the Chegem River, Labar
danSu River valley
* Samples from collection of S.N. Bubnov (IGEM RAS) and
** Samples from the collection of V.Yu. Gerasimov (IGEM RAS).
Note: Bi is biotite, cPx is clinopyroxene, oPx is orthopyroxene, Pl is plagioclase, Amf is amphibole, Q is quartz, and Ol is olivine.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 47
the isotope dilution method on a Micromass Sector 54
(Great Britain) mass spectrometer. The accuracy in
determination of Sr and Nd isotope composition were
controlled by systematic analyses of SRM987 and La
Jolla standards, respectively. The measurement errors
in
87
Sr/
86
Sr
and
143
Nd/
144
Nd
ratios were not greater than
1 and 0.3%, respectively.
The Pb isotope composition was studied using the
MCICPMS method. Samples (from 50 to 150 mg in
weight) were preliminarily treated with 2M
HNO
3
for
30 minutes at a temperature of
60–80
°
С
in order to
remove the adsorbed Pb. Samples were dissolved in a mix
ture of concentrated HF +
HNO
3
mixed in the proportion
3 : 1 in Teflon vessels. Pb was separated using anion
exchanged columns (0.1 ml of BioRadAG1x8 resin) and 1
M HBr medium. Isotope Pb ratios were measured on a
MCICPMS NEPTUNE (ThermoFinnigan, Ger
many) mass spectrometer. Mass discrimination effect was
corrected for reference
205
Tl/
203
Tl
= 2.3889, which was
measured in each mass spectrum (scan) strictly simulta
neously with measurement of Pb isotope ratios. Analyzed
Pb solutions were preliminarily spiked with standard
Tl sample. The measurement uncertainties in Pb isotope
ratios in experiments (
±
2SE
) vary from 0.006 to 0.024%,
being for most samples less than total analytical errors
(
±
2SD
). The latter were estimated from the results of
Table 2. Results of K–Ar isotope dating of groundmass of Quaternary lavas of Elbrus volcano
Sample no. Potassium
, %
±σ
40
Ar
rad
,
ng/g
±σ
40
Ar
air
,
in sample Age, ka
±
2
σ
Lavas of phase III (Late Neopleistocene–Holocene)
393 2.61
±
0.03
not detected ≥
99.9
≤
30
Az21 2.56
±
0.03 0.0025
±
0.0016 99.6 15
±
15
Az22 2.50
±
0.03 0.0062
±
0.0018 99.0 35
±
20
Az23 2.38
±
0.03 0.0033
±
0.0012 99.2 20
±
10
Az26 2.83
±
0.03 0.0052
±
0.0024 96.8 25
±
15
Az28 2.88
±
0.03 0.0059
±
0.0028 98.4 30
±
20
Az29 2.78
±
0.03
not detected ≥
99.9
≤
30
E4 2.80
±
0.03 0.0068
±
0.0017 95.4 35
±
20
E5 2.68
±
0.03 0.0017
±
0.0017 98.7 10
±
10
E7 3.28
±
0.03
not detected ≥
99.9
≤
30
E8 2.96
±
0.03 0.0052
±
0.0015 98.9 25
±
15
Ml37 3.04
±
0.03 0.007
±
0.003 94.9 30
±
25
Lavas of phase II (Late Neopleistocene)
1993Ea 2.97
±
0.03 0.0193
±
0.0013 93.1 95
±
15
Az19 3.12
±
0.03 0.0200
±
0.0015 89.9 90
±
15
Az30 2.86
±
0.03 0.0206
±
0.0023 93.7 105
±
20
Az31 3.07
±
0.03 0.0238
±
0.0024 90.3 110
±
20
E2 3.00
±
0.03 0.0232
±
0.0016 87.1 110
±
15
Ml36 3.23
±
0.03 0.015
±
0.002 85.3 70
±
20
Ml43 3.30
±
0.03 0.016
±
0.002 83.8 70
±
20
Lavas of phase I (Middle Neopleistocene) phase
Az25 3.54
±
0.04 0.0414
±
0.0016 84.4 170
±
15
E1 3.00
±
0.03 0.0426
±
0.0011 79.3 205
±
10
E3 3.00
±
0.03 0.0377
±
0.0017 89.4 180
±
15
E6 3.79
±
0.04 0.0258
±
0.0019 88.2 100
±
15
Bt6 3.05
±
0.04 0.046
±
0.003 64.9 220
±
30
Bt17 2.86
±
0.04 0.045
±
0.003 75.2 225
±
30
Ml41 3.14
±
0.04 0.047
±
0.003 93.4 215
±
30

48
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
62 parallel analyses of SRM 981 standard and
accounted for:
206
Pb/
204
Pb –
±
0.016%;
207
Pb/
204
Pb –
±
0.016%;
208
Pb/
204
Pb –
±
0.018%;
207
Pb/
206
Pb –
0.005%, and
208
Pb/
206
Pb –
±
0.009%
[21].
Obtained isotope data were recalculated using
decay constants recommended by the International
Geochronological Subcommission [22].
RESULTS
Petrographic Review of Lavas of Elbrus Volcano
Detailed mineralogical–petrological studies of the
Elbrus lavas were conducted by many researchers and
published in numerous papers and monographs [2, 4,
5, 23, 24, and others]. In this relation, we present only
brief petrographic characteristics of volcanic samples
studied by us.
Dacitic lavas from Elbrus volcano display mainly
porphyritic texture. The phenocrysts are represented
by ubiquitous plagioclase (from andesine–labradorite
to oligoclase), orthopyroxene, and biotite, and less
common amphibole and quartz. In some varieties,
quartz is rimmed by clinopyroxene. Microphenocrysts
of clinopyroxene (augite) are extremely rare.
Two generations of phenocrysts are usually present
in lavas. Phenocrysts of the first generation are repre
sented by crushed and melted crystals and fragments
of plagioclase, quartz, orthopyroxene, hornblende,
and variably deformed and opacitized biotite laths.
Plagioclase in composition corresponds to oligoclase–
andesine; orthopyroxene is hypersthene and biotite is
characterized by Fe/Fe + Mg = 45–50%. Amphiboles
are ascribed to intermediate members of common
hornblende–magnesioriebeckite join, but more often
their composition is close to common hornblende.
Phenocrysts of the second generation are unde
formed and often represented by zoned crystals. Pla
gioclases typically have labradorite cores and andesine
rims. Fe mole fraction in zoned orthopyroxenes varies
from 18–20% in the core to 25–28% in their margins.
The groundmass typically shows a fluidal microstruc
ture and consists of partially or completely devitrified
volcanic glass. Its composition is dominated by plagio
clase with less abundant orthopyroxene (bronzite),
biotite and subordinate clinopyroxene (augite).
Results of K–Ar Dating of Lavas of Elbrus Volcano
In order to compile the chronological scale of the
Elbrus eruptions, we carried out the isotopegeochro
nological study of known reference sections on its
southern slope (Azau valley), as well as two sections on
the northern (upper reaches of the Malka River) and
western (upper reaches of the BiitikTebe River)
slopes. The assignment of these sections to published
geological maps of the volcano [4–6] gives grounds to
state that dated samples were taken from lava flows of
different age and different phases of Elbrus activity:
from early to, supposedly, Holocene ones. Thus, the
isotope dating of samples from our collection provides
an opportunity to unravel the entire evolution of
Elbrus and to determine the periods of its maximal
activity and time of last eruptions.
According to our first field observations, the lava
flows on the southern slope of Elbrus volcano (Azau
valley) form at least four independent scarps (hori
zons), which mark the eruptions of different age [11].
Based on subsequent detailed geological mapping in
the Azau valley, we specified the section of the south
ern slope of Elbrus and, correspondingly, distin
guished five stratigraphic horizons of lavas (Fig. 2). In
the area of Terskol settlement, the lavas of the lowest
horizon rest on the Paleozoic metamorphic schists
100–150 m above the present erosion level of the Azau
I
II III
IV
V
PZ
II
I
V
I
II
VII II
V
III
PZ
PZ
I
II
V
PZ
PZ
PZ
Fig. 2.
Panoramic view of the southern slope of Elbrus volcano with deciphering lava flows of different age. PZ are rocks of the
Paleozoic basement.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 49
River valley and are traced to the river head, where
they plunge beneath the Bol’shoi Azau glacier. Lava
flows of adjacent horizons in some sections are sepa
rated by glacial deposits, suggesting fairly long gaps
between lava eruptions.
The most complete section of lava flows on the
southern slope of Elbrus volcano is represented in the
Terskol Range (watershed of the Terskol and Gara
bashi rivers). In this area, the rocks of all five horizons
are subsequently exposed in the earthroad lacet from
the Terskol settlement to the astronomic observatory
and Ice Base ruins. The rocks of the lowest horizon (I)
armor range slopes at altitudes from 2350–2400 to
2700–2900 m. Small lava remain of this horizon pre
served also on the left bank of the Terskol River, near
AchiSu mineral spring. The volcanics of horizon II
are exposed at altitudes from 2700 to 3000 m, forming
cliffs with radial columnar joints in the Volchii Vorota
locality and bench that gives way to the Devichii Kosy
waterfall. The lavas of horizon III compose the range
on the Levaya and Pravaya Garabashi river watershed
and are traced to the area of stake 105. The horizon IV
is represented by lava flow of the Terskol Range and
forms a plateau at altitudes of 3000–3200 m from stake
105 to the Terskol astronomical observatory. In the
area of Ice Base and stake 105, the lavas of horizon IV
lie on the volcanics of stage III and are overlain by the
youngest lavas of horizon V, which make up the so
called “Ice Base” flow approximately 300 m thick
(Fig. 2).
To the west, on the watershed of the Garabashi and
Malaya Azau rivers (Fig. 2), the lavas of horizon IV are
absent. In the area of cable way, lava horizon III
pinches out, while the older volcanics are overlain by
extended (approximately 10 km) young lava flows of
the uppermost horizon V, which takes its origin on the
Eastern summit and descends southward (Pastukhov
Rocks–Refuge of 11—Garabashi station—Mir Sta
tion) up to the modern Azau River bed. In the area of
the Azau cable way station, in the river bed, it forms a
tongue approximately 1 km long, from which the
Malaya Azau River plunges as waterfall cascade.
In the Bol’shaya and Malaya Azau rivers watershed,
the lava flows of different age make up practically a
single bench a few hundreds meters high. Our observa
tions showed that this section contains lavas of hori
zons I, II, and V (Fig. 2).
Based on field observations, results of deciphering
the satellite images and analysis of previously pub
lished maps [4–6], we compiled geological map (scale
1 : 50000) of the southern slope of Elbrus (Fig. 3),
which specifies the geological structure of this part of
the volcano. The localities of studied samples together
with obtained K–Ar isotope datings were plotted on
the map (Fig. 3) and geological scheme of the volcano
(Fig. 1).
On the western slope of Elbrus volcano, the section
of its lava flows was sampled in the upper reaches of the
BiitikTebe River (Fig. 1). According to our data, the
Elbrus lavas in this area rest on the Pale ozoic basement
and pyroclastic rocks (tuffs, ignimbrites, tufflavas) of
PaleoElbrus volcano of Eopleistocene (approxi
mately 800–700 ka) age [1, 10]. According to geomor
phological data, the lava flows of the BiitikTebe River
valley, in general, are ascribed to the single, oldest
phase of Elbrus evolution [6], but some data [5] sug
gest that they represent the eruption products of sepa
rate subsidiary volcanic edifice. We sampled volcanics
from the lower and upper parts of the section.
In the upper reaches of the Malka River (northern
slope of Elbrus Volcano), the volcanic section was
studied in the Birdzhalysu River above the Zhelasu
thermal mineral waters (Fig. 1). In this area, the oldest
Elbrus lavas are represented by flows in the
Birdzhalysu River head, where they lie immediately
on the Paleozoic basement, while the younger lavas
are made up of the flow overlaying the Kyzylkol River
(left tributary of the Malka River) near the Zhelasu
springs (“Sultan waterfall” flow) and stratigraphically
latest flows of the Irakhiktyuz locality. Wide and rela
tively gentle northern slopes of Elbrus allowed the
youngest lavas of Elbrus to be distributed over abun
dant areas, while the absence of deeply incised river
valleys and largescale propagation of moraine and
river deposits in this area strongly prevent the direct
observation of stratigraphic relationships between
lavas of different age.
Table 2 presents the results of K–Ar dating of the
rocks of Elbrus volcano. Obtained age values are dis
tinctly clustered into three groups: 205–170, 110–90,
and less than 35 ka.
Dates on lavas from the BiitikTebe River valley
(western slope) and the Birdzhalysu River head
(northern slope), as well as all rocks of the lowermost
horizon I from the southern slope of Elbrus, except for
the date on sample E6, fall in the time range of 205–
170 ka corresponding to the Middle Neopleistocene.
The age values obtained for five lava samples of hori
zons II–IV from the southern slope, as well as two
samples from the “Sultan waterfall” lava flow in the
Malka River heads are practically indistinguishable
within measurement error and plot in a range of 110–
70 ka (Late Neopleistocene). The most numerous
group of samples (12) characterizing the rocks of the
youngest horizon V of the Azau valley and flow of the
Irakhiktyuz locality on the northern slope of Elbrus
define ages less than 35 ka, which indicates that these
volcanics were formed at the end of the Late Neopleis
tocene or Holocene. Ultralow contents of radiogenic
40
Ar
in the youngest lavas of Elbrus (0.005–0.002 ng/g)
and, as a result, generally low ratio
40
Ar
rad
/
40
Ar
tot
determine the relatively high uncertainty in K–Ar dat
ings. However, the content of radiogenic
40
Аr
in total
40
Ar for some samples reaches 2–3%, which makes
this analyzed material favorable (for corresponding
age range) for K–Ar dating and indicates, with confi
dence, the young age (<35 ka) of these rocks. Note
worthy is the good agreement between dates obtained

50
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
Fig. 3.
Geological map of the southern slope of Elbrus volcano (compiled by V.A. Lebedev based on field observations, decipher
ing satellite images, and works of Koronovskii and Gazeev [4, 5], and largescale geological maps of the area [6]). (
1
) glaciers and
firn fields; (
2
) Quaternary sedimentary rocks; (
3
) Late Neopleistocene–Holocene lavas; (
4
) Late Neopleistocene lavas; (
5
) Mid
dle Neopleistocene lavas; (
6
) Paleozoic basement; and (
7
) sampling localities and obtained K–Ar data.
25
E8
5642
Western
summit
Eastern
summit
Pastukhov Rocks
⭐
30
Refuge of 11 E7
IRIK GLACIER
TERSKOL GLACIER
GARABASHI GLACIER
Ice Base
393E
Az22
St. Mir
⭐
30
St. Garabashi
MALYI AZAU GLACIER
BOL’SHOI AZAU GLACIER
Mt. Azaubashi
CHIPERAZAU GLACIER
CHEGETKARA GLACIER
KYUNNYUMAK GLACIER
NENSKRA GLACIER
Tersko R.
Sarykol R.
observ.
Terskol
1993Ea
Az31
E2
St. Azau
hotel Azau
E3
stake 105
Az25
M. Azau R.
St. Staryi Krugozor
E 5
Az26
Az21
Az28Az29
Terskol
E1
min spr.
Baksan R.
hotel Cheget
Dongu
Mt. Azaugitche–Cheget–Karabashi
Mt. Donguzorunbashi
Mt. AzauCheget–Karabashi
20
25
10
100 15
30
90 110 95
105
205
⭐
30
170
35
Az19
GEORGIA
123 4 56 7
90
1 km
Elbrus volcano
180
Garabashi R.
110
35
Az23
scient. base
E 6
Azau R.
E 4
Az30
zorun R.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 51
for samples from the same lava flows (for instance, for
samples Ml36 and Ml43 of the “Sultan waterfall”
flow).
Thus, the stratigraphic sequence of the lava flows
on the slopes of Elbrus volcano, which was inferred by
previous researchers and locally specified by us during
mapping, is well consistent with obtained geochrono
logical data. The exception is the lava sample E6 from
the lowermost horizon of the Azau valley with K–Ar
age of
100
±
15
ka corresponding to the Late–Neo
pleistocene age group (Table 2). This can be related to
a partial loss of radiogenic Ar from the rocks due to
thermal impact, because this sample was taken in the
roof of the lava horizon immediately overlain by the
Late Pleistocene–Holocene lavas (sample E5).
Of special interest is the date on lavas from the
Western summit of Elbrus (sample E8). According to
geomorphological data, the Western partially
destroyed cone is older than the eastern one [5, and
others]. This is consistent with the absence of volca
nics with age less than 200 ka in the sections of the
western slope. It could be expected that the rock from
the Western summit should fall in the Late Neopleis
tocene or Middle–Neopleistocene age group; how
ever, its age of
25
±
15
ka unambiguously suggests its
very young age. The obtained age of sample E8 is
consistent with two interpretations. First, both craters
(Western and Eastern Elbrus) were formed and oper
ated in the Late Neopleistocene–Holocene, while
lavas with age approximately 200 and 100 ka were
erupted earlier from their precursor. The second ver
sion suggests that activity of Western Elbrus lasted dur
ing a long time (from 200 to less than 35 ka), whereas
Eastern Elbrus formed in the Late Neopleistocene–
Holocene. Thus, the question of activity periods of the
Elbrus cones remains open and requires more detailed
geological and isotope–geochronological study.
Extended young lava flow descending in the Azau
River valley from the Eastern summit of Elbrus along the
cable way (flow of the Malaya Azau River) was sampled in
most detail. Its samples yielded a series of K–Ar dates,
which differ in errors, but definitely indicate their forma
tion age of less than 35 ka (Table 2, samples Az21,
Az22, Az23, Az28, and Az29). As was mentioned
above, Bogatikov et al. [12, and others] report several
14
C
dates, four of which are regarded to be related to the
eruption of lava flows of Elbrus and its satellite TashTebe
volcano, as well as outburst of pyroclastic material during
explosive eruptions. One of these dates (
4270
±
40
years)
was obtained for coals from soil horizon resting on the
considered lava flow in the area of Azau meadow. It is
known that much time is required to form soils on the
stony surface under highmountainous conditions, and
the gap between lava eruptions and formation of overlay
ing soils can reach many thousands years. Therefore, the
discussed radiocarbon date, as most other dates, presum
ably determines only the upper age limit of lava erup
tions. Considering our K–Ar and published radiocarbon
data [12], the age of the youngest volcanic rocks of Elbrus
volcano lies within 35–4 ka.
Isotope–geochronological data obtained for lavas
sections from different slopes of Elbrus make it possible
to reconstruct the time scale of its evolution. First, the
results of K–Ar dating indicate that the volcanic activity
of Elbrus lasted no more than 250 ka. In our opinion,
three age ranges distinguished by us from K–Ar dates for
Elbrus volcanics correspond to three phases of its maxi
mal activity. Phase I (Middle Neopleistocene) with an
age of 225–170 ka was responsible for the formation of
the lowermost horizon lavas of the Azau valley (southern
slope), most lava flows of the BiitikTebe River (western
slope) and the Birdzhalysu River head (northern slope).
Phase II (Late–Neopleistocene, 110–70 ka) produced
second, third, and fourth lava horizons of the Azau
valley on the southern slope of Elbrus. On the north
ern slope, this phase was responsible for the formation
of the “Sultan waterfall” flow, which overlaid the
Kyzylkol River and stopped at the area of the Zhelasu
mineral springs. The possible eruption centers of the
Middle Neopleistocene and Late Neopleistocene
lavas could be the western, older cone of Elbrus; how
ever, in light of obtained data, we can suggest that for
mation of volcanics of phases I and II was related to
the other presently destroyed edifice, which was later
replaced by two modern cones.
The products of phase III (Late Neopleistocene–
Holocene), with an age less than 35 ka, were noted on
the southern and northern slopes of Elbrus. In the
Azau valley, they form lava flows near the Irik glacier,
in the Ice base area, and the watersheds of the Gara
bashi–Malaya Azau and Malaya Azau–Bol’shaya
Azau rivers. In the area of the Malka River basin, they
cover the most part of the northern Elbrus slope,
where older flows and basement rocks are fragmen
tarily exposed: lavas usually form separate remains,
while Paleozoic granites and metamorphic rocks are
exposed in the individual erosion windows. According
to obtained K–Ar and published radiocarbon dates
[12], the age of phase III volcanics is no more than
35 ka and can be as young as a few thousand years.
Note that the studied Late Neopleistocene–Holocene
lavas of Elbrus, in general, differ in lower K content in
the groundmass as compared to the rocks of early
eruptive phases of the volcano.
Thus, the volcano is characterized by discrete vol
canic activity (eruptions) for the last 250–200 ka, with
periods of maximal intensity of its eruptions in the
Middle Neopleistocene (225–170 ka), middle Late
Neopleistocene (110–70 ka), and terminal Late Neo
pleistocene–Holocene (less than 35 ka).
Petrogeochemical Characteristics of Elbrus Lavas
Petrology, geochemistry, and mineral composition
of Elbrus lavas, as well as xenoliths and melt inclusions
in them were considered in numerous works and dis
sertations [2, 4, 23, and others]. The main conclusion

52
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
of these studies was the hybrid mixed mantle–crustal
origin of the Elbrus volcanics. In this case, the compo
sitional specifics of these lavas resulted from contami
nation of primary mantle melts by crustal material or
mixing of mantle and crustal melts [2].
New isotopegeochronological data made it possi
ble to decipher the chronology of the Elbrus evolution
and correspondingly, to trace the compositional evo
lution of volcanics with time, and, taking into account
the hybrid origin of studied lavas, to estimate the pos
sible change in role of mantle and crustal sources at
various phases of volcanic activity.
Table 3 reports the results of determination of
petrogenic oxides and some trace elements in the stud
ied Elbrus lavas. As follows from this table and Fig. 4,
the Elbrus rocks correspond to dacites and less com
mon rhyodacites. They contain 65.2–70.4%
SiO
2
and
6.4–7.9%
K
2
O + Na
2
O
at 2.7–3.9%
К
2
О
and are
ascribed to calcalkaline petrochemical series. The
MiddleNeopleistocene rocks, in general, differ in
elevated contents of silica and alkalis as compared to
the younger lavas (Fig. 4). Note that Elbrus lavas are
characterized by relations Cr > Ni > Co (Table 3),
which is typical of hightemperature and deepseated
magmatic rocks. According to classification diagram
SiO
2
–K
2
O
(Fig. 5), the studied volcanics of phases I
and II belong to highK type, while rocks of phase III
belong to high and moderatepotassium types.
The contents of most petrogenic oxides (except for
SiO
2
and
K
2
O
) and trace elements (except for Rb)
show no systematic variations versus age of the studied
rocks (Table 3). The Mg index of the Elbrus lavas var
ies within a wide range (Mg# from 0.25 to 0.49), show
ing no correlation with rock age. Thus, the products of
the Elbrus volcanic activity have fairly homogeneous
chemical composition, which revealed no significant
variations during the history of volcano eruptions.
In the petrogenetic diagrams Nb–Y and Rb–(Y +
Nb) [25], the data points of the Elbrus lavas are close
to the average compositions of I and Stype granites
[26], but, in the second diagram (Fig. 6), they define a
linear trend from the field of syncollisional granitoids
toward the volcanic arc granites. In the
(Na
2
O + K
2
O–
CaO)–SiO
2
discriminant diagram [27], the data points
of all studied Elbrus rocks plot in the field of I and
Stype granites. In the diagrams
K
2
O–Na
2
O
and Ba–
Rb–Sr, they are plotted in the field of Itype granites.
To sum up, the petrochemical characteristics of the
studied dacite lavas of Elbrus are close to those of syn
and postcollisional Itype granitoids, which is consis
tent with the mixed mantle–crustal nature of parental
magmas of the volcano [2]. The evolution of the
Elbrus melts during more than 200ka period of its
activity was expressed in a minor increase of mafic
components against the complementary decrease in K
and Rb contents. The noted trend may be related to
the increase of mantle contribution in the volcanics at
the late phases of the Elbrus activity.
Sr, Nd, and Pb Isotope Composition of Elbrus Lavas
We studied Sr–Nd isotope systematics in all dated
lava samples of Elbrus volcano (Tables 4, 5). The
Elbrus rocks, in general, are characterized by elevated
contents of Sr (210–335 ppm) and Rb (140–280 ppm).
Note that Sr variations show no correlation with lava age
(Table 4). In contrast, the Rb contents in the volcanics of
phase III strongly differ from those in phases I and II
(140–155 and 155–220 ppm, respectively). Thus, the
Fig. 4.
TAS diagram for the studied rocks of Elbrus vol
cano. Boxes are rocks of phase I, triangles are rocks of
phase II, and crosses are rocks of phase III.
Fig. 5.
Diagram SiO
2
–K
2
O for the studied rocks of Elbrus
volcano. Symbols are shown in Fig. 4.
9
8
7
6
578726862585654 76747064 6660
Rhyolites
Trachydacites
Dacites
Andesites
Trachyandesites
Basaltic andesites
SiO
2
, wt %
K
2
O + Na
2
O, wt %
4
3
2
1
071676665 706968
SiO
2
, wt %
K
2
O, wt %
Highpotassium type
Moderatepotassium type
Lowpotassium type
I
II
III

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 53
products of Late Neopleistocene–Holocene volcanic
activity are depleted in Rb as compared to dacites of
early phases, which agrees with the recognized trend
of increasing mafic components at decrease of K con
tent in lavas during Elbrus evolution. The exception is
two dacite samples of phase III: E7 (Irik flow) and
MI37 (lavas of the Irakhiktyuz locality), which have
notably higher Rb contents (200–280 ppm) as com
pared to coeval volcanics. This is, presumably,
explained by the presence of a significant amount of
biotite (up to 5%) in these lavas. Notable differences in
Rb contents between dacitic lavas of different age,
respectively, determine significant differences in their
Rb/Sr ratios: 1.6–1.2 in the rocks of phase III (except
for samples MI37 and E7) against 2.9–1.5 in the
rocks of phases I and II (Table 4).
The Sm and Nd contents in the studied lavas of
Elbrus volcano are 3–7 and 16–42 ppm, respectively
(Table 5). In general, the products of individual phases
of Elbrus activity do not show any enrichment or
depletion in these elements. The
147
Sm/
144
Nd
isotope
ratios (Table 5) in the rocks fall in the range of 0.099–
0.141, which is typical of either crustal rocks or prod
ucts of withinplate mantle magmatism. Note that
similar values of this isotope ratio were previously
found in the young rocks from other Caucasian
regions: subalkali basalts from lava rivers of the Dzha
vakheti highland in South Georgia (0.105–0.137)
[28], subalkali basalts of Central Georgia (0.101–
0.114) [29], and Pliocene Dzhimara granitoid massif
in the Kazbek neovolcanic area (0.109–0.144) [30].
The range of initial
87
Sr/
86
Sr
and
143
Nd/
144
Nd
ratios
in the rocks of Elbrus volcano is fairly wide: 0.70535–
0.70636 and 0.51252–0.51268 (or from –2.3 to +0.8
in
Є
Nd
units), respectively (Tables 4, 5). Of special
interest are significant differences in Sr and Nd iso
tope composition between lavas of different volcanic
phases. Of all the studied Elbrus rocks, the most radio
genic Sr (0.70587–0.70636) and least radiogenic Nd
(
Є
Nd
from –2.2 to –0.9) compositions were deter
mined in dacites of phase I. The Late Neopleistocene
lavas of phase II are characterized by intermediate
ratios of
87
Sr/
86
Sr
and
143
Nd/
144
Nd
(0.70569–0.70578
and
Є
Nd
from –1.8 to –1.1), while the Late Neopleis
tocene–Holocene lavas of phase III have the least
radiogenic Sr (0.70535–0.70559) and most radiogenic
Nd (
Є
Nd
from –1.4 to + 0.8). The exceptions are
already discussed dacite samples of phase III (MI37
and E7), whose Sr and Nd isotope compositions
(
87
Sr/
86
Sr
= 0.70587–0.70602 and
Є
Nd
from –2.3 to
–1.4) are not close to those of simultaneous lavas, but
approximate rocks of phase I (Table 4). Thus, the lavas
of Elbrus volcano show systematic trends in Sr and Nd
isotope composition with time: decrease in
87
Sr/
86
Sr
and increase in
143
Nd/
144
Nd
(Fig. 7).
Note that data points of the Elbrus lavas in the dia
gram
87
Sr/
86
Sr
–1/Sr show a rough linear correlation
between the content and isotope composition of Sr
(Fig. 8). Such a correlation and systematic change in
Sr isotope composition in lavas can be interpreted as
evidence of the mixing of Sr from two different
sources, mantle and crustal, in the parental magmas.
This conclusion is consistent with petrogeochemical
features of the Elbrus rocks, which, as was mentioned
above, can also result from their mixed mantle–crustal
genesis. The contribution of crustal and mantle
sources in the genesis of Quaternary rocks of the
Elbrus volcanic center was previously demonstrated
also by studying the He isotope composition in gases
of the thermal springs from the Elbrus region, in which
3
He/
4
He
accounted for 2.2–8.7
×
10
–6
[31].
Based on previous studies of young mafic volcanic
rocks from the Greater and Lesser Caucasus, we
assumed that practically all Late Cenozoic magmatic
rocks of this region were derived from a single mantle,
possibly lowermantle source [29, 30], which, in terms
of its petrogeochemical characteristics, was close to
hypothetical “Common” source [32]. This source was
termed “Caucasus” and has the following isotope–
geochemical characteristics:
87
Sr/
86
Sr = 0.7041
±
0.0001, Є
Nd
= +
4.1
±
0.2
at
147
Sm/
144
Nd
= 0.10–0.12
[30]. The chemical composition of primary melts pro
duced by this source possibly corresponds to K–Na
subalkali basalts. Among Neogene–Quaternary rocks
of the Greater Caucasus, only basic rocks of the Chia
tura center in the Central Georgian area with an age of
6.3–6.1 Ma (chemical composition is shown in
Table 3, Rb–Sr and Sm–Nd data are given in Tables 4
and 5) are close to “Common” reservoir in their Sr and
Nd signatures [29].
In the Sr–Nd isotope diagram (Figs. 9, 10), the
data points of Elbrus lavas are approximated by the
line close to the mixing curve between the “Caucasus”
depleted mantle and crustal reservoirs. It is interesting
that Sr–Nd data points of other young magmatic
rocks from the Elbrus neovolcanic area are also con
2000
1000
100
10
10012000100010 Y + Nb,
ppm
Rb,
ppm
Withinplate
Oceanridge granites
Syncollisional granitoids
Volcanicarc granites
Fig. 6.
Discriminant Rb–(Y + Nb) diagram [25]. Symbols
are shown in Fig. 4.
granitoid

54
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
fined to this line. These are data that were obtained in
this work (Tables 4, 5) or published previously: ignim
brites and tuffs from the upper reaches of the Malka
River (
87
Sr/
86
Sr – 0.70751, Є
Nd
= –2.8, [2]), ignim
brites from the upper reaches of the BiitikTebe River
(
87
Sr/
86
Sr – 0.70638, Є
Nd
= –2.2 [2]), trachybasaltic
andesites from the Tyzyl River (
87
Sr/
86
Sr – 0.70503–
0.70507, Є
Nd
from +0.4 to +0.6 [3]), trachybasaltic
andesites from the Surkh and Krandukh volcanoes
(
87
Sr/
86
Sr – 0.70460–0.70464, Є
Nd
from +1.8 to +2.2
[3]), granites of the Eldjurtu Massif (
87
Sr/
86
Sr
=
0.7068–0.7071 [33, 34];
Є
Nd
= –2.4, Table 5), and
granitoids of the Caucasian Mineral Water region
(
87
Sr/
86
Sr – 0.7083–0.7086, [35]; Є
Nd
from –3 to –4,
Table 5). Thus, the position of data points of the Elbrus
lavas, as well as other Neogene–Quaternary magmatic
rocks from the Elbrus area, in the Sr–Nd diagram pre
sumably testifies to the common genesis of the Late
Cenozoic magmatic rocks of this region of the Greater
Caucasus. Their parental melts were likely the prod
ucts of interaction between contrasting mantle melts
and crustal material.
Figures 9 and 10 show data points of young rocks
from the Elbrus volcanic area relative to fields of hypothet
ical mantle reservoirs EMI, EMII, U, COMMON,
DMM [32, 36, 37], and inferred average composition of
the lower crust (LC) [38]. It is seen that the parental mag
mas of the Neogene–Quaternary complexes were derived
by mixing of the mantle source “Caucasus” with source
having sufficiently high
87
Sr/
86
Sr
(>0.710) and low
143
Nd/
144
Nd (Є
Nd
< –5)
. Obviously, the aforementioned
mantle reservoirs cannot serve as this source. The
model data point of the average composition of the
lower continental crust [38] is also far removed from
mixing hyperbole and cannot be its termination. In
addition, the Sr and Nd isotopic compositions of
eclogites from the Chegem River valley (50 km east of
Elbrus volcano), which could be considered as possi
ble example of lower crustal rocks at the Greater Cau
casus that were exhumed to surface conditions owing
to tectonic movements and Alpine orogenesis in sepa
rate blocks, have extremely high present
Є
Nd
+ 7.6 and
low present
87
Sr/
86
Sr
= 0.7035 (Sample 72, Tables 4
and 5) values. Based on chemical composition of these
rocks (Table 3) and isotope characteristics, these
eclogites were formed during the Late Paleozoic
regional metamorphism of ancient oceanic basalts. In
the Sr–Nd diagrams (Figs. 9 and 10) the data point of
the eclogites is plotted far from mixing curve defined
by data points of young magmatic rocks of the Elbrus
area, being confined to MORB field. The possible
contribution of these rocks in the genesis of the young
est lavas of Elbrus seems to be ambiguous. This, pre
1
0
1
2
3 0.7070.705
87
Sr/
86
Sr
Є
Nd
0.706
Fig. 7.
Vari a t ions o f
87
Sr/
86
Sr
versus
Є
Nd
in the studied
dacitic lavas of different age from Elbrus volcano. Symbols
are shown in Fig. 4.
0.708
0.707
0.706
0.705
0.704 0.00440.00400.00360.00320.0028 0.0048
1/Sr
87
Sr/
86
Sr
Fig. 8.
Diagram
87
Sr/
86
Sr–1/Sr
for dacitic lavas of Elbrus volcano. Symbols are shown in Fig. 4.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 55
Table 3. Chemical composition of studied lava samples from Elbrus volcano, Late Miocene basalts from the Chiatura center of the Central Georgian neovolcanic area, and
Paleozoic rocks from the crystalline basement of Elbrus volcano
Sample
SiO
2
TiO
2
Al
2
O
3
Fe
2
O
3
MnO MgO CaO Na
2
OK
2
OP
2
O
5
SCrScVCoNiCuZnRbSrYZrNbBaPbClGa
wt % ppm
Lavas of Elbrus volcano
Bt
6 68.08 0.75 15.35 4.07 0.06 0.91 3.41 3.89 3.19 0.24 0.05 47 4 n.d. 9 40 10 63 135 368 11 259 n.d. 391 7 n.d. 19
Bt
17 70.11 0.68 14.74 3.22 0.05 0.54 2.99 3.91 3.51 0.23 0.01 50 9 n.d. 11 40 10 70 142 378 11 205 n.d. 403 10 n.d. 16
Ml
36 68.00 0.74 15.33 4.11 0.06 0.85 3.28 4.00 3.39 0.22 0.02 32 8 n.d. 9 22 16 66 159 389 6 202 n.d. 492 21 n.d. 14
MI37
68.45 0.74 14.98 4.29 0.06 0.88 3.27 3.77 3.35 0.20 0.01 40 7 n.d. 8 28 16 69 160 359 8 234 n.d. 461 23 n.d. 13
Ml
41 67.62 0.75 15.44 4.16 0.06 1.02 3.44 4.02 3.24 0.24 0.01 41 9 n.d. 10 32 4 63 148 392 8 204 n.d. 481 27 n.d. 15
Ml
43 68.20 0.73 15.27 4.06 0.06 0.86 3.26 3.92 3.41 0.23 0.01 30 7 n.d. 8 22 18 67 158 365 15 233 n.d. 381 27 n.d. 20
393 67.32 0.79 15.51 4.28 0.06 1.23 3.54 4.10 2.96 0.21 n.d. 30 9 n.d. 9 n.d. n.d. n.d. 139 330 n.d. n.d. n.d. 491 n.d. n.d. n.d.
1993
Ea
67.33 0.79 15.37 4.16 0.06 1.47 3.53 3.77 3.27 0.25 n.d. 42 8 n.d. 9 n.d. n.d. n.d. 136 340 n.d. n.d. n.d. 549 n.d. n.d. n.d.
Ea
19 66.13 0.69 16.21 4.04 0.09 1.86 3.43 4.00 3.31 0.23 n.d. 75 8 n.d. 9 31 16 51 151 413 6 204 n.d. 414 29 n.d. 9
Ea
21 65.64 0.79 16.11 4.64 0.10 1.84 4.10 3.77 2.77 0.23 n.d. 78 9 n.d. 10 45 11 67 114 433 8 200 n.d. 341 24 n.d. 16
Ea
22 66.27 0.76 15.71 4.55 0.09 1.82 3.59 4.27 2.72 0.23 n.d. 75 8 n.d. 9 43 9 72 126 345 9 204 n.d. 414 11 n.d. 15
Ea
23 65.18 0.83 15.79 5.05 0.09 2.10 3.89 3.99 2.84 0.24 n.d. 42 9 n.d. 10 41 14 67 114 355 8 210 n.d. 277 18 n.d. 8
Ea
25 68.11 0.76 14.62 4.11 0.08 1.60 2.59 3.99 3.90 0.24 n.d. 29 7 n.d. 8 32 8 63 177 283 7 186 n.d. 433 48 n.d. 14
Ea
26 67.33 0.74 15.79 4.09 0.08 1.77 3.50 3.27 3.20 0.24 n.d. 43 8 n.d. 9 40 10 60 139 375 9 215 n.d. 462 30 n.d. 14
Ea
28 66.42 0.73 15.74 4.31 0.09 1.82 3.50 4.00 3.13 0.25 n.d. 73 9 n.d. 10 43 11 66 138 392 6 186 n.d. 435 20 n.d. 16
Ea
29 67.22 0.76 15.22 4.24 0.07 1.89 3.48 3.63 3.24 0.25 n.d. 33 7 n.d. 8 35 14 58 144 375 9 213 n.d. 297 13 n.d. 14
Ea
30 66.58 0.79 15.33 4.63 0.09 2.25 3.65 3.27 3.18 0.24 n.d. 82 8 n.d. 9 44 12 62 140 398 7 206 n.d. 342 42 n.d. 10
Ea
31 66.59 0.76 15.57 4.42 0.09 1.92 3.49 3.53 3.40 0.23 n.d. 65 7 n.d. 9 32 9 56 143 390 7 221 n.d. 337 40 n.d. 10
E1 70.10 0.70 13.94 3.85 0.06 0.93 2.93 3.63 3.54 0.30 0.02 22 9 68 6 15 22 61 152 298 21 233 17 470 28 96 22
E2 67.87 0.63 15.53 3.77 0.06 1.05 3.77 3.91 3.15 0.24 0.03 28 10 58 7 13 18 62 132 384 21 209 13 466 23 320 23

56
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
Table 3.
(Contd.)
Sample
SiO
2
TiO
2
Al
2
O
3
Fe
2
O
3
MnO MgO CaO Na
2
OK
2
OP
2
O
5
SCrScVCoNiCuZnRbSrYZrNbBaPbClGa
wt % ppm
E3 68.25 0.63 15.52 3.61 0.06 0.85 3.68 3.87 3.26 0.25 0.02 20 8 68 5 14 20 55 139 373 21 215 15 401 25 300 25
E4 68.35 0.66 14.93 3.99 0.06 1.19 3.64 3.90 3.00 0.26 0.02 26 13 65 10 13 29 67 125 335 20 221 12 421 27 149 21
E5 69.14 0.68 14.43 3.89 0.06 1.08 3.40 3.88 3.16 0.26 0.02 20 12 61 8 13 23 62 128 316 20 226 12 425 28 126 22
E6 69.38 0.62 14.61 3.43 0.06 1.09 3.03 3.89 3.59 0.29 0.02 18 7 62 8 13 21 57 162 296 21 213 15 440 28 228 20
E7 70.41 0.57 14.36 3.13 0.05 1.18 2.78 3.56 3.72 0.21 0.03 14 9 49 8 12 11 53 166 279 19 191 14 510 30 337 22
E8 69.78 0.62 14.44 3.51 0.05 1.09 3.04 3.87 3.36 0.21 0.02 17 8 51 8 11 23 58 147 288 20 206 12 415 23 289 19
Late Miocene basalts of the Chiatura center of the Central Georgian area
ZG
04 46.81 1.68 15.44 10.43 0.17 8.70 9.96 4.04 2.01 0.72 0.03 371 27 164 44 125 48 76 24 777 22 162 25 420 6 199 19
ZG
05 46.11 1.71 14.52 10.81 0.17 10.33 10.23 3.52 1.87 0.71 0.02 399 26 152 52 131 60 83 23 839 25 169 29 467 6 331 13
ZG
06 46.45 1.72 14.44 10.66 0.17 10.62 10.15 3.19 1.87 0.68 0.04 380 22 156 53 136 52 83 23 812 23 163 26 409 6 283 19
Rocks of the Paleozoic basement of Elbrus volcano
CH21 51.64 1.22 25.60 10.32 0.17 3.49 0.55 1.57 5.27 0.08 0.08 108 17 134 30 47 63 153 220 127 39 182 22 692 55 58 43
Az5 66.57 0.52 15.20 8.03 0.15 2.49 1.17 1.96 3.74 0.12 0.05 49 14 70 15 34 34 116 160 180 20 82 11 398 41 57 20
Az3 71.49 0.24 14.35 2.21 0.03 0.81 1.69 2.26 6.63 0.24 0.04 8 8 8 1 7 9 52 195 60 23 94 13 706 46 152 18
Az4 70.12 0.52 15.14 3.61 0.03 1.29 1.71 2.74 4.65 0.18 0.02 4 6 28 3 9 5 77 152 170 18 199 14 719 17 74 23
SH28 73.14 0.33 13.18 3.48 0.08 0.75 1.25 2.49 5.14 0.13 0.02 11 11 38 6 10 2 45 103 155 37 175 11 1118 43 147 15
AR403 46.72 1.41 24.28 15.54 0.45 7.57 1.84 1.58 0.22 0.01 0.38 180 26 168 73 248 169 81 220 117 66 257 16 36 11 210 29
72 46.75 1.02 13.17 12.63 0.17 8.74 13.68 3.58 0.11 0.05 0.09 328 n.d. 230 20 121 63 197 2 321 16 51 4 78 3 n.d. n.d.
Notes: Analyses were made by A.I. Yakushev and T.M. Marchenko in IGEM RAS by the XRF method on a Philips PW 2400 spectrometer. Results of measurements of rockforming oxides
were recalculated to 100%; n.d. is not determined.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 57
sumably, implies an insignificant role of the lower
crustal rocks in the formation of parental melts for the
rocks of the Elbrus neovolcanic area.
The upper crustal Paleozoic granitemetamorphic
complexes of the Greater Caucasus seem to be the
most suitable crustal source which took part in the
genesis of young magmatic rocks of the Elbrus area
together with mantle source “Caucasus.” We studied
six samples of metamorphic schists, gneisses, and
granitoids from the collection of V.Yu. Gerasimov
(IGEM RAS). These samples were taken in the vicin
ity of Elbrus volcano and represent the rocks of its
basement (Table 1). The chemical composition of the
studied Paleozoic rocks is listed in Table 3, while their
Sr–Nd isotope characteristics are shown in Tables 4
Fig. 9.
Sr–Nd isotope–correlation diagram for lavas of Elbrus volcano and other young rocks from the Elbrus neovolcanic area.
The curve shows the mixing between depleted mantle of the “Caucasus” source and hypothetical average composition of the
upper granite–metamorphic crust of the Greater Caucasus (PZ). Data on the rocks from the Elbrus area were taken from this
work, as well as from works [2, 33–35]; compositions of mantle sources and lower continental crust were taken from [32, 36, 37].
Asterisks denote inferred end members of mantle and upper crustal reservoirs, which took part in the genesis of the Neogene–
Quaternary magmatic rocks of the Elbrus neovolcanic area.
Fig. 10.
Sr–Nd diagram illustrating approximate proportions of mantle and crustal material in the parental magmas of lavas of
Elbrus volcano and other magmatic complexes of the Elbrus neovolcanic area. Data on the rocks from the Elbrus area were taken
from this work, as well as from papers [2, 33–35]; data on mantle sources and lower continental crust were taken from [32, 36,
37]. See symbols in Fig. 4.
15
10
5
0
5
10
15 0.750.730.720.710.70 0.74
87
Sr/
86
Sr
Є
Nd
DMM
Eclogites of the Greater Caucasus
С
HIMU
UM
EM I
EM II
LC
Paleozoic granitoids and metamorphic rocks of the Greater Caucasus
Mixing curve
“Caucasus”
Symbols
Ignimbrites and tuffs of the upper riches of Malka River
Ignimbrites and tuffs of the BiitikTebe River
Trachybasaltic andesites of the Tyzyl R.
Trachybasaltic andesite of the Surkh and Krandukh volcanoes
Granitoids of CMW area
Granites of the Eldjurtu Massif
Lavas from Elbrus volcano
Subalkali basalts from the Central Georgian area
Paleozoic granitoids and metamorphic
rocks of the Greater Caucasus
10
8
6
4
6
8
10 0.7100.7080.7040.7030.702 0.709
87
Sr/
86
Sr
C
Nd
DMM
Eclogites of the Greater Caucasus
HIMU
U
EM I
EM II
LC
0.705 0.706 0.707
4
2
0
2
0.4“Caucasus”
0.5“Caucasus”
0.6“Caucasus”
0.7“Caucasus”
0.8“Caucasus”
0.9“Caucasus”
Common
Central Georgia
Krandukh
Tyzyl
Elbrus Eldjurtu Massif
Granitoids of
Ignimbrites and tuffs of
the Malka R.
Ignimbrites of
the BiitikTebe R.
CMW area
“Caucasus”

58
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
Table 4. Results of Rb–Sr isotope study of the Late Quaternary lavas from Elbrus volcano, Late Miocene subalkali basalts from
the Chiatura center of the Central Georgian neovolcanic area, and Paleozoic rocks of the crystalline basement of Elbrus volcano
Sample
Rb,
ppm
Sr,
ppm
87
Rb/
86
Sr
±
2
σ
87
Sr/
86
Sr
±
2
σ
Lavas of phase III (Late Neopleistocene–Holocene) of Elbrus volcano activity
393 150 280 1.54
±
1 0.705586
±
15
Az
22 140 300 1.33
±
1 0.705542
±
15
Az
26 155 320 1.39
±
1 0.705558
±
15
Az
28 160 310 1.49
±
10.705536
±
15
Az
29 155 350 1.30
±
10.705559
±
15
Az
23 140 335 1.19
±
1 0.705403
±
15
Az
21 145 330 1.29
±
1 0.705353
±
15
E4 155 285 1.55
±
1 0.705527
±
15
E5 150 320 1.37
±
1 0.705503
±
15
E7 200 210 2.73
±
20.706016
±
15
E8 180 245 2.09
±
2 0.705566
±
21
MI
37 276 298 2.71
±
2 0.705868
±
14
Lavas of phase II (Late Neopleistocene) of Elbrus volcano activity
Az
19 175 280 1.84
±
1 0.705776
±
15
Az
30 155 310 1.47
±
1 0.705768
±
15
Az
31 170 290 1.71
±
1 0.705777
±
15
1993Ea 175 295 1.72
±
10.705773
±
15
E2 170 285 1.69
±
1 0.705760
±
15
Ml
36 197 299 1.94
±
2 0.705690
±
13
Ml
43 214 284 2.21
±
2 0.705754
±
16
Lavas of phase I (Middle Neopleistocene) of Elbrus volcano activity
Az
25 205 215 2.75
±
20.705836
±
15
E1 175 290 1.77
±
1 0.705815
±
15
E3 175 280 1.80
±
1 0.705811
±
15
E6 220 220 2.91
±
20.705826
±
15
Bt
6 168 280 1.75
±
2 0.706358
±
14
Bt
17 176 333 1.55
±
2 0.706112
±
14
Ml
41 209 279 2.18
±
20.705904
±
15
Late Miocene subalkali basalts from the Chiatura center of the Central Georgian neovolcanic area
ZG
4 36 1100 0.0997
±
4 0.704110
±
14
ZG
5 25 790 0.0917
±
40.704065
±
14
ZG
6 35 1050 0.0963
±
4 0.704041
±
14
Rocks of the Paleozoic basement of Elbrus volcano
CH
21 250 124 5.84
±
3 0.750502
±
15
AZ5 175 180 2.83
±
1 0.736470
±
15
AZ3 210 63 9.68
±
40.750036
±
15
AZ4 160 180 2.57
±
10.720128
±
15
SH28 110 170 1.92
±
1 0.720782
±
15
AR403 220 110 5.67
±
2 0.710324
±
15
72 1.1 480 0.0070
±
6 0.703505
±
15
Notes: Analyzed material of lavas of Elbrus volcano and Late Miocene subalkali basalts from the Chiatura center of the Central Georgian
neovolcanic area was represented by groundmass; other materials were represented by wholerock samples. Table lists measured values
of 87Sr/86Sr ratio (for young volcanic rocks, they correspond to initial values).

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 59
Table 5. Results of Sm–Nd isotope study of the Late Quaternary lavas from Elbrus volcano, young magmatic rocks from some
of neovolcanic areas of the Greater Caucasus, and Paleozoic rocks from crystalline basement of Elbrus volcano
Sample
Sm,
ppm
Nd,
ppm
147
Sm/
144
Nd
±
2
σ
143
Nd/
144
Nd
±
2
σ
Є
Nd
Lavas of phase III (Late Neopleistocene–Holocene) of Elbrus volcano activity
393 4.5 26 0.1037
±
30.512580
±
10 –1.1
Az22 5.4 31 0.1042
±
2 0.512601
±
11 –0.7
Az26 2.6 16 0.1032
±
30.512580
±
12 –1.1
Az28 5.4 31 0.1032
±
3 0.512565
±
10 –1.4
Az29 5.3 31 0.1028
±
30.512570
±
10 –1.3
Az23 6.8 38 0.1077
±
3 0.512679
±
10 +0.8
Az21 4.9 28 0.1053
±
3 0.512645
±
11 +0.1
E–4 5.0 29 0.1034
±
30.512615
±
10 –0.4
E5 5.0 31 0.1036
±
30.512618
±
10 –0.4
E7 5.0 28 0.1020
±
3 0.512565
±
10 –1.4
E8 5.7 33 0.1027
±
30.512586
±
10 –1.0
Ml37 n.d. n.d. n.d. 0.512518
±
14 –2.3
Lavas of Phase II (Late Neopleistocene) of Elbrus volcano activity
Az19MI37 7.1 42 0.1018
±
3 0.512565
±
10 –1.4
Az30 5.7 33 0.1031
±
30.512568
±
10 –1.4
Az31 5.7 33 0.1032
±
30.512579
±
10 –1.1
1993Ea 5.3 32 0.0986
±
3 0.512569
±
10 –1.3
E2 4.0 24 0.1033
±
30.512568
±
10 –1.4
Ml36 n.d. n.d. n.d. 0.512565
±
9–1.4
Ml43 n.d n.d n.d 0.512544
±
11 –1.8
Lavas of phase I (Middle Neopleistocene) of Elbrus volcano activity
Az25 5.2 32 0.0991
±
30.512590
±
10 –0.9
E1 5.7 33 0.1044
±
30.512587
±
10 –1.0
E3 5.0 25 0.1411
±
40.512567
±
10 –1.4
E6 6.0 34 0.0980
±
3 0.512575
±
10 –1.2
Bt6 n.d. n.d. n.d. 0.512530
±
9–2.1
Bt17 n.d n.d n.d. 0.512527
±
19 –2.2
Ml41 n.d. n.d. n.d. 0.512551
±
8–1.7
Late Miocene subalkali basalts from the Central Georgian area
ZG
4 6.3 33 0.1144
±
3 0.512832
±
10 +3.8
ZG
5 6.5 35 0.1116
±
20.512827
±
10 +3.7
ZG
6 6.5 34 0.1136
±
20.512848
±
10 +4.1
Granitoids of the Caucasian Mineral Water region
MW
5 5.5 33 0.0995
±
1 0.512437
±
8–3.9
MW
8 1.4 10 0.0885
±
1 0.512472
±
7–3.2
MW
9 7.8 48 0.0979
±
4 0.512457
±
7–3.5
Granites of the El’djurtu Massif
EL1 4.2 22 0.1160
±
1 0.512515
±
6–2.4
Rocks of the Paleozoic basement of Elbrus volcano
CH21 16.7 101 0.0995
±
2 0.511899
±
10 –14.4
AZ5 2.4 11.8 0.1210
±
2 0.511934
±
10 –13.7
AZ3 3.2 11.5 0.1678
±
2 0.512173
±
10 –9.1
AZ4 4.0 19.7 0.1209
±
2 0.512102
±
20 –10.5
SH28 4.1 18.1 0.1360
±
2 0.512134
±
10 –9.8
AR403 6.6 42 0.0956
±
2 0.511969
±
10 –13.1
72 1.4 4.2 0.1960
±
40.513026
±
14 +7.6
Notes: Analyzes of lavas of Elbrus volcano and Late Miocene subalkali basalts from the Chiatura center of the Central Georgian neovolcanic
area were made for groundmass; for other samples, we analyzed whole rock; n.d is not determined. The values of in table cor
respond to presentday values (for Neogene–Quaternary volcanics they within error indistinguishable from initial
ЄNd
ЄNd
T).

60
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
and 5. As was expected for heterogeneous strongly dif
ferentiated upper crust, the granitoids and metamor
phic rocks show significant variations in contents of
major and trace elements. These rocks contain 60–
180 ppm Sr (averaging 140 ppm), 110–250 ppm Rb
(approximately 200 ppm, on average), 2–17 ppm Sm
(approximately 5 ppm, on average), 10–100 ppm Nd
(approximately 30 ppm, on average), at
87
Rb/
86
Sr
from
2 to 10 and
147
Sm/
144
Nd
from 0.095 to 0.168 (Tables 4,
5). The Paleozoic rocks also show strong differences in
the presentday Sr isotope composition (
87
Sr/
86
Sr
from
0.710 to 0.750) at less significant variations in Nd iso
topic composition (
Є
Nd
from –9 to –14) (Tables 4, 5).
Due to geologically insignificant time scales (less than
200 ka), the measured presentday
87
Sr/
86
Sr
and
Є
Nd
values in the granites and metamorphic schists and
gneisses should be practically identical to those at the
time of formation of Elbrus lavas.
Our data and available limited results of Sr–Nd
isotope analyses of Paleozoic metamorphic rocks and
granitoids confirm a considerable heterogeneity of the
Greater Caucasus upper crust, which prevent accurate
calculation of its average Sr–Nd characteristics.
Therefore, we can approximately estimate only the
contribution of crustal and mantle sources for differ
ent age lavas of the Elbrus volcano and other young
magmatic rocks of the Elbrus neovolcanic area.
As was mentioned above, the Sr–Nd isotope char
acteristics of the “Caucasus” mantle source calculated
by us using data on subalkali basalts from several neo
volcanic areas of the Greater and Lesser Caucasus to
be as follows:
87
Sr/
86
Sr – 0.7041
±
0.0001, Є
Nd
– +4.1
±
0.2
[28–30]. Based on chemical composition of basic
rocks of the Chiatura district of the Central Georgian
area of the Greater Caucasus, which as we believe is
fairly close to that of primary mantle magmas of the
“Caucasus” source, they contain ~800 ppm Sr (Table 3)
and 35 ppm Nd (Table 5). The Sr and Nd isotopic
composition and contents of these elements in the
hypothetical upper crustal reservoir of the Greater
Caucasus were calculated as average of six studied
samples of Paleozoic rocks from the Elbrus volcano
basement: Sr – 140 ppm, Nd – 30 ppm,
87
Sr/
86
Sr –
0.730, Є
Nd
– –12 (Tables 3–5). The mixing hyperbole
calculated between these two end members (Figs 9,
10), in general, well approximates the position of data
points of the young magmatic rocks from the Elbrus
neovolcanic area in the Sr–Nd isotope correlation
diagram. The close Nd contents in both the sources
define a sufficiently gentle mixing curve; significant
(by more than a factor of five) differences in Sr content
between primary magmas of “Caucasus” source and
upper crustal rocks of the Greater Caucasus define
asymmetric distribution of data points along mixing
hyperbole: they are accumulated in its left portion
(Fig. 9).
According to this estimate, the maximal contribu
tion of upper crustal rocks (approximately 50–60%) in
the genesis of parental magmas of the young rocks
within the Elbrus area is noted for the Late Miocene
granitoids of the Caucasian Mineral Water region,
while basic lavas of the Tyzyl flow and Krandukh vol
cano demonstrate the minimal (less than 20%) crustal
contribution. Dacite lavas of Elbrus are characterized
by intermediate proportions of crustal and mantle
materials; the role of the “Caucasus” source in the
petrogenesis of volcano magmas significantly
increased with time (from ~60 to ~80%) from prod
ucts of phase I to phases II and III (Fig. 10). Thus,
modeling conducted shows that during Elbrus evolu
tion over last 200 ka, the role of mantle material in its
products sequentially increased, which was reflected
in the systematic change of isotopegeochemical char
acteristics, increase of mafic components, and
decrease of
K
2
O
content in the lavas. Hybrid melts
were formed at the initial phase of Elbrus evolution by
mixing of mantle and crustal sources and possibly pre
served in the intermediate magmatic chambers. Dur
ing phases II and III, they interacted with new por
tions of deepseated mantle magmas. As was noted by
Bubnov [2], these “injections” of hightemperature
basic melt into chamber with cooling dacitic magma
served as specific trigger to the resumption of volcanic
eruptions.
Of special interest in this context are data obtained
for samples MI37 and E7, which are in conflict with
revealed evolution trends in petrogeochemical and
isotope characteristics of the parental magmas of
Elbrus with time. These rocks formed at the last III
phase of volcanic activity, but their Sr–Nd character
istics are close to those of phase I dacites (Figs. 7–10).
One can suggest that volcanic rocks of the Irik flow
(sample E7) and lavas of the Irakhiktyuz locality
(sample MI37) were formed from relict melt, which
preserved since the initial phase of the Elbrus activity
in some separate magmatic chambers and erupted on
the surface with a delay of 200 ka. The ejections of
these relict lavas at phase III can be caused, for
instance, by tectonic movements leading to the
squeezing of “relict” melt or its heating against
resumed eruptive activity. It cannot also be excluded
that the Irik flow and lavas of the Irakhiktyuz locality
were formed at the initial stage of the formation of the
Eastern Elbrus cone, when opening of new magma
channels could lead to the extrusion of “relict” mag
mas from the peripheral chambers of Elbrus. Note that
samples MI37 and E7 represent essentially porphy
ritic rocks (30–50% phenocrysts), which indicates the
long residence time of their parental magmas in inter
mediate chambers beneath the volcano.
Pb isotope composition was studied in the dacites
from the southern slope section, which compose lava
flows of all three phases of volcanic activity. Among
eight analyzed samples, there is also dacite from the
Western summit of Elbrus (Table 6). As for Sr and Nd,
the Pb isotope composition was also determined in the
subalkali basalts of the Chiatura center of the Central
Georgian volcanic area (two samples), which charac

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 61
terize the inferred “Caucasus” mantle source, in the
Paleozoic granitoids and metamorphic schists taken
near the volcano (four samples), as well as in the
amphibolized eclogite (sample 72) as representative
of the lower crustal material of the Greater Caucasus.
The results are presented in Table 6.
The Pb content in the studied Elbrus dacites (Table 6)
varies in a narrow range (23–30 ppm), which, in gen
eral, is higher than that in the felsic and intermediate
volcanic rocks commonly [39]. Variations of Pb isoto
pic ratios are within 18.631–18.671 for
206
Pb/
204
Pb,
15
.649–15.660 for
207
Pb/
204
Pb
, and 38.811–38.847 for
208
Pb/
204
Pb
. Relative differences in end values are,
respectively, 0.21, 0.07, and 0.09%.
The following circumstances must be taken into
account in estimating these variations. Most of the
presently available Pb isotopic data on magmatic rocks
(like on other natural objects) were obtained using
thermoionization mass spectrometry (TIMS) with
measurement uncertainties depending on mass differ
ence of lead isotopes that are involved in measured iso
tope ratios, and according to estimates presented in
many works account for approximately 0.1, 015, and
0.2% for ratios
206
Pb/
204
Pb,
207
Pb/
204
Pb
, and
208
Pb/
204
Pb
,
respectively. The analysis of previously published lead
isotopegeochemical data shows that the error in iso
topic ratios analyzed by TIMS in low Pb (
<10
–3
%
)
rocks and minerals in many cases is really higher than
indicated above. This is related to the uncontrolled
effect of isotope fractionation in sample as well as
Table 6. Pb isotope composition of lavas from the southern slope of Elbrus volcano, Late Miocene basalts from the Chiatura
center of the Central Georgian neovolcanic area, and Paleozoic rocks from the crystalline basement of Elbrus volcano
Sample
Pb,
ppm
206
Pb/
204
Pb
±
2SE
207
Pb/
204
Pb
±
2SE
208
Pb/
204
Pb
±
2SE
Lavas from the southern slope of Elbrus volcano
E1 28 18.6516
±
20 15.6595
±
20 38.8455
±
48
E2 23 18.6705
±
24 15.6536
±
22 38.8253
±
56
E3 25 18.6427
±
20 15.6512
±
20 38.8200
±
40
E4 27 18.6322
±
40 15.6487
±
20 38.8108
±
54
E5 28 18.6654
±
12 15.6558
±
10 38.8173
±
26
E6 28 18.6338
±
34 15.6520
±
30 38.8263
±
80
E7 30 18.6464
±
26 15.6534
±
22 38.8297
±
60
E8 23 18.6430
±
32 15.6555
±
28 38.8305
±
62
Late Miocene subalkali basalts from the Central Georgian area
ZG
4 6 18.7239
±
16 15.6181
±
16 38.7837
±
42
ZG
6 6 18.7178
±
36 15.6149
±
32 38.7718
±
78
Paleozoic rocks from the crystalline basement of Elbrus volcano
CH2 55 18.5446
±
16 15.6910
±
12 39.1191
±
38
AZ3 46 18.4604
±
16 15.6939
±
16 38.7015
±
42
AZ4 17 18.4404
±
10 15.6934
±
10 38.7515
±
28
AZ5 41 18.2485
±
12 15.6663
±
12 38.7121
±
36
72 3 18.2620
±
40 15.5953
±
36 38.0035
±
92
Note: Groundmass was analyzed in lavas of Elbrus volcano and Late Miocene basalts of the Chiatura center of the Central Georgian neovol
canic area; for other samples, we analyzed wholerock samples; Pb contents were determined by the XRF method (Table 3).

62
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
other difficulties accompanying TIMS isotope analy
sis of lead. In any case, geochemically significant dif
ferences between Pb isotopic ratios obtained by TIMS
analyses should exceed the indicated minimal level of
analytical errors. The new MCICPMS method,
methodical and geochemical aspects of which with
respect to studying Pb isotope composition were con
sidered in our recent work [21], strongly changed this
situation. Since this method is much more accurate
than TIMS analysis, the difference between geochem
ically significant and insignificant variations of
206
Pb/
204
Pb,
207
Pb/
204
Pb
, and
208
Pb/
204
Pb
ratios mea
sured by the MCICPMS method decreased by
almost an order of magnitude: up to 0.02%. Note that
variations observed in the Elbrus lavas are 3–10 times
higher than this value and, at the same time, lie within
analytical errors typical for the traditional TIMS
method. The considered data on the Elbrus dacites,
for the first time in isotope geology, provide insight
into the true scale of Pb isotope variations in cogenetic
series of compositionally close volcanic rocks and an
opportunity to correctly compare it with isotope com
position of other magmatic rocks.
Unlike
87
Sr/
86
Sr
ratios, Pb isotope ratio in the
Elbrus dacites shows no correlation with their age. At
the same time, there is negative correlation between
208
Pb/
204
Pb
and
Na
2
O/K
2
O
(Fig. 11). In the
206
Pb/
204
Pb–
207
Pb/
204
Pb
diagram (Fig. 12), the data
points of Elbrus dacites form a field, which lies well
above the average Stacey and Kramers model crust
evolution line (
µ
2
= 9.74), left of its presentday zero
point. Correspondingly, the studied volcanic samples
are characterized by high (crustal) values of
µ
2
within
9.84–9.88 and model Pb–Pb age within a range from
77 to 106 Ma.
Significant variations (up to 1.6% with respect to
206
Pb/
204
Pb
) in Pb isotope composition between stud
ied Paleozoic basement rocks could be explained by
different composition and, possibly, age of these rocks.
As may be seen in Table 6, the Elbrus lavas strongly
differ in Pb isotope composition from other studied
rocks: the Late Miocene subalkali basalts of the Cen
tral Georgian neovolcanic area and Paleozoic crystal
line rocks that compose the base of the volcano. How
ever, the character of these differences and, corre
spondingly, the mutual position of data points in the
diagrams
206
Pb/
204
Pb–
207
Pb/
204
Pb
and
206
Pb/
204
Pb–
208
Pb/
204
Pb
(Fig. 12) and
87
Sr/
86
Sr–
206
Pb/
204
Pb
(Fig. 13), where the compact field of the Elbrus dac
ites is plotted between subalkali basalts representing
the “Caucasus” mantle source and the Paleozoic
granitoids and metamorphic rocks, confirms previous
inferences from Sr–Nd data that the parental magmas
of Elbrus were formed by a mixing of mantle and crustal
materials. The mixed character of Pb isotope composi
tion of the Elbrus lavas is consistent with the fact that the
total Pb content in the dacites (7–48 ppm) is intermedi
ate between its content in subalkali basalts (6 ppm) and
ancient crystalline rocks (10–55 ppm) (Table 3). The cal
culation of lead isotope data, like those based on Sr–Nd
isotopy, yields the mixing of mantle and crustal materials
in the proportion of 55 and 45% for Elbrus lavas. These
estimates slightly differ toward higher upper crustal con
tribution in the parental magmas of Elbrus as compared
to calculations based on Sr–Nd data (60–80% of “Cau
casus” source and 40–20% material of Paleozoic grani
toids and metamorphic rocks). However, it should be
noted that the accuracy for mixing proportions between
mantle and crustal sources in the Elbrus lavas estimated
using lead isotopic compositions is strongly limited by
low Pb contents in the subalkali basalts of the central
Georgian area (6 ppm) determined with fairly high error.
The assimilation of Pb from upper crustal rocks
during their interaction with mantle magmas mainly
and first of all could be realized by Pb mobilization
from K feldspars as main Pb carriers in these rocks.
From this viewpoint, the participation of upper crustal
rocks similar to rocks that compose the Elbrus base
ment is supported by the correlation between
208
Pb/
204
Pb
and
Na
2
O/K
2
O
ratios (Fig. 11). Potassium
feldspars, as carriers of common lead, were sources of
all its isotopes. However, additional source of
206
Pb
and
207
Pb
, especially the former, during assimilation of
upper crustal rocks of granite composition could be
radiogenic lead from both Ubearing accessory miner
als and layered uranium compounds.
The overestimated model Pb–Pb ages (i.e., age cal
culated from
207
Pb/
206
Pb
) of Elbrus dacites are possibly
related to the presence of relatively older common
lead, which was derived, for instance, from feldspars in
the basement crystalline rocks during their assimila
tion by mantle magmas.
1.3
1.2
1.1
1.0
0.938.81 38.82 38.83 38.84 38.85
Na
2
O/K
2
O
208
Pb/
204
Pb
Fig. 11.
208
Pb/
204
Pb
–
Na
2
O/K
2
O
in the studied lavas of
Elbrus volcano.

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 63
The Pb isotope composition of the studied eclogite
(Sample 72) sharply differs from those in the Elbrus
dacites (Fig. 12). The data point of this rock in the iso
tope diagram (Figs. 12, 13) is plotted far from data
points of Elbrus lavas, “Caucasus” source, and hypo
thetical mixing curves. Like Sr–Nd isotope data, this
may indicate insignificant contribution of lower
crustal rocks of the Greater Caucasus in the petrogen
esis of the Elbrus lavas.
Thus, the obtained lead isotopegeochemical data are
well consistent with concepts envisaging the upper
crustal rocks as one of sources of dacitic lavas of Elbrus.
It should also be noted that lead isotope character
istics of subalkali basalts of the Central Georgian area,
in particular, their negative Pb–Pb model age and
lowered
µ
2
indicate the presence of isotopically anom
alous lead in the “Caucasus” reservoir. Published data
confirm that this is, in general, typical of hypothetical
Fig. 12.
Diagram
206
Pb/
204
Pb–
207
Pb/
204
Pb
and
206
Pb/
204
Pb–
208
Pb/
204
Pb
for the studied lavas from Elbrus volcano, subalkali
basalts from the Central Georgian area, and Paleozoic metamorphic and magmatic complexes of the Greater Caucasus. Pb iso
tope data on different mantle and crustal reservoirs were taken from the work [32].
39.2
39.0
38.8
38.6
38.4
38.2
38.0
15.70
15.68
15.66
15.64
15.62
15.60
15.58 19.618.618.418.218.0
17.817.6 18.8 19.0 19.2 19.4
COMMON
LCC
SK = (
ω
2
= 36.84)
0
100
200
300
208
Pb/
204
Pb
Mixing curve
SK(
μ
2
= 9.74)
COMMON
0
100
200
300
400
500
Dacites of Elbrus
Subalkali basalts of the Central Georgian
area (“Caucasus”)
Paleozoic granitoids and
metamorphic rocks (upper crust
of the Greater Caucasus)
Eclogites (sample 72)
SK is the average evolutionary curve
according to Stacey–
206
Pb/
204
Pb
207
Pb/2
04
Pb
Kramers model

64
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
lowermantle sources such as “Common,” “FOZO,”
and others [for instance, 32].
CONCLUSIONS
Complex isotope–geochronological and petro
geochemical studies indicate that the Elbrus volcano
experienced a long and complex evolution, while the
products of its activity are hybrid rocks that were derived
by the mixing of deepseated mantle source with upper
continental crust of the Greater Caucasus.
The K–Ar dating of the Elbrus lavas from reference sec
tions on its southern, western, and northern slopes indicates
that its volcanic activity lasted 250–250 ka totally. Isotope–
geochronological data show that Elbrus is not a continu
ously active volcano, but its evolution was intermittent:
largescale eruptions were separated by long (approxi
mately 50 ka) quiescence periods. Obtained K–Ar dates
made it possible to distinguish three phases of maximal
intensity of the Elbrus eruption: Middle Neopleistocene
(225–170 ka), Late Neopleistocene (110–70 ka), and Late
Neopleistocene–Holocene (less than 35 ka). No eruptions
presumably occurred during “quiescence” periods, while
the volcano was dormant or revealed only insignificant
explosive eruptions and postmagmatic activity.
The discrete character of the eruptive activity of
Elbrus and its insufficiently long lifetime (for instance,
Kazbek volcano erupted during last 450–400 ka [10,
40]), first, leave open the problem of cessation of its late
phase and entire volcanic activity, and, second, suggest
the probability of resumption of eruptions in this region.
Note that a few available radiocarbon ages constrain the
upper age limit of the youngest lava flows of Elbrus to a
few thousand years.
Late Neopleistocene–Holocene dates obtained for
lava flows from the Eastern summit of the volcano and
rocks that immediately compose its Western summit sug
gest that Elbrus acquired its presentday appearance only
at the late phase of its eruption. In the Middle and begin
ning of the Late Neopleistocene, this area was likely
occupied by volcanic cone, which was later destroyed by
largescale glacial erosion; the edifice of the modern
Western summit and later, in vicinity, Eastern summit
were formed in this area less than 35 ka. Distinctly
observed asymmetry of the Western cone related to
decomposition of its southwestern part indicates a high
intensity of the tectonic movements in this area, which,
Fig. 13.
Diagram
87
Sr/
86
Sr –
206
Pb/
204
Pb
for the studied lavas from Elbrus volcano, subalkali basalts from the Central Georgian
area and Paleozoic rocks from the metamorphic and magmatic complexes of the Greater Caucasus. Data on Pb isotope compo
sition for different mantle and crustal reservoirs were taken from the work [32].
23
22
21
20
19
18
17 0.70 0.71 0.72 0.73 0.74 0.75
87
Sr/
86
Sr
206
Pb/
204
Pb
HIMU
COMMON
“CAUCASUS”
EM I
EM II
Mixing curve
LCC
DMM
Upper crust of the Greater Caucasus
Symbols
Dacites of Elbrus
Eclogites (sample 72)

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 65
during the period of maximal duration of tens of thou
sands of years, destroyed a significant segment of large
volcanic edifice.
The results of petrogeochemical and Sr–Nd–Pb iso
tope–geochemical studies indicate a mixed mantle–
crustal origin of the Elbrus lavas. According to our data,
hybrid parental magmas of the volcano were formed
owing to mixing and/or contamination of deepseated
mantle melts by material of the Paleozoic upper crust of
the Greater Caucasus.
Mantle reservoir that contributed in the genesis of
Elbrus volcano, as most other Neogene–Quaternary
magmatic rocks of the Greater and Lesser Caucasus [30],
was presumably represented by lowermantle source,
which in composition and isotopic characteristics is close
to the “Common” source [32]. Based on previous results
[28–30] and data from this work, we suggest that primary
melts generated by this lower mantle source (termed as
“Caucasus”) in composition correspond to K–Na subal
kali basalts and is characterized by the following isotope
characteristics:
87
Sr/
86
Sr = 0.7041
±
0.0001, Є
Nd
=
4.1
±
0.2,
147
Sm/
144
Nd = 0.105–0.114,
206
Pb/
204
Pb = 18.72,
207
Pb/
204
Pb
= 15.62, and
208
Pb/
204
Pb
= 38.78. Note that,
in addition to low
147
Sm/
144
Nd
isotope ratios typical of
crustal rocks or products of withinplate magmatism, the
“Caucasus” source is characterized by isotopically
anomalous Pb composition corresponding to the nega
tive Pb–Pb model ages. Our estimates indicate that the
“Caucasus” generated melts that were significantly
enriched in Sr (up to 800 ppm), and had elevated Nd
contents (approximately 30 ppm) and lowered Pb (less
than 10 ppm).
Most data points of young magmatic rocks of the
Elbrus area, including the Elbrus volcano, are plotted on
the Sr–Nd mixing hyperbole calculated between lower
mantle “Caucasus” source and averaged estimated com
position of the Paleozoic upper crust of the Greater Cau
casus. This fact may indicate a common genesis of all
Neogene–Quaternary magmatic rocks of this area.
Approximate proportions of contribution of mantle and
crustal sources in the genesis of parental rocks of Elbrus
account for from 1.5 to 4 depending on rock age.
Of special importance is the fact that proportions of
crustal and mantle material in the parental magmas of
the volcano gradually changed toward the increasing
role of the latter: from ~60% at the Middle–Neopleis
tocene phase to ~80% at the Late Neopleistocene–
Holocene phase. This phenomenon is also supported
by a general decrease in contents of
SiO
2
, K
2
O
, and Rb
in the lavas of late eruptions. This indicates that the
activity of deepseated source did not decrease, but
increased during evolution. The role of crustal melts or
contamination steadily decreased. We suggest that
resumption of volcanic activity during II and III
phases was related to the influx of hightemperature
mantle melt in the intermediate magmatic chambers
at upper crustal levels, which already contained unso
lidified dacitic magmas that have not been erupted on
the surface during previous pulses of activity. Owing to
the mixing of new portions of subalkali basaltic melt
with moderately acid magma, the latter became
increasingly mafic and overheated dacites that rapidly
ascended to the surface.
To sum up, the deciphering carried out of the evo
lution of the volcano and the discrete character
revealed of its activity together with the absence of
signs of cessation of Late Neopleistocene–Holocene
phase III, increasing role of deepseated mantle
source established from isotope–geochemical data, as
well as numerous geophysical evidence for the pres
ence of subsurface magmatic chambers with liquid
magma beneath the volcano [17, and others], the pres
ence of fumaroles in the summit part and thermal
springs along the periphery of volcano—all these facts
suggest that Elbrus is a potentially active volcano with
high degree of possibility of resumption of volcanic
activity. The three most possible scenarios can be pro
posed for continuation of the eruptive activity of the
volcano. The first scenario suggests the continuation
of phase III with eruption in the nearest time (hun
dreds–thousands years) of flows of hightemperature
mobile dacitic lavas similar in composition to those
erupted for the last 35 ka. The second scenario sug
gests that cessation of phase III and onset of a long
period of quiescence in the nearest few tens of thou
sands of years, during which Elbrus will be in a state of
quiescence, with no signs of any intense eruptive activ
ity. The resumption of eruptions at a new hypothetical
phase of activity will be, presumably, caused by geo
tectonic reasons, in particular, by change in the con
vergence regime between the Arabian and Eurasian
plates accompanied by increasing compression within
the Greater Caucasus zone, which could lead to the
opening of new or regeneration of old magmatic chan
nels beneath the volcano. The third, catastrophic sce
nario, suggests the formation of large collapse caldera
and decomposition of volcanic edifice owing to the
presence of numerous discharged chambers beneath
it. This process would be accompanied by largescale
pyroclastic ejecta, as it took place at the Middle–Late
Pliocene boundary during formation of the Upper
Chegem Caldera situated 50 km east of Elbrus.
ACKNOWLEDGMENTS
The work was supported by the Russian Founda
tion for Basic Research (project no. 060564763a)
and by the Priority Research Program No. 16 of the
Presidium of the Russian Academy of Sciences.
REFERENCES
1. I. V. Chernyshev, V. A. Lebedev, S. N. Bubnov, et al.,
“Stages of Magmatic Activity in the Elbrus Volcanic
Center (Greater Caucasus): Evidence from Isotope
Geochronological Data,” Dokl. Akad. Nauk
380
(3),
384–389 (2001) [Dokl. Earth Sci.
380
, 848–852
(2001)].

66
GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
LEBEDEV et al.
2. S. N. Bubnov, Extended Abstract of Candidate’s Dis
sertation in Geology and Mineralogy (IGEM RAN,
Moscow, 2003).
3. V. A. Lebedev, S. N. Bubnov, I. V. Chernyshev, et al.,
“Basic Magmatism in the Geological History of the
Elbrus Neovolcanic Area, Greater Caucasus: Evidence
from K–Ar and Sr–Nd Isotope Data,” Dokl. Akad.
Nauk
406
(1), 78–82 (2006) [Dokl. Earth Sci.
406
, 37–
40 (2006)].
4. V. M. Gazeev, Extended Abstract of Candidate’s Dis
sertation in Geology and Mineralogy (IGEM RAN,
Moscow, 2003).
5. N. V. Koronovskii, “Geological Structure and Evolu
tion of Elbrus Volcano,” in
Elbrus Glaciation
(Mosk.
Gos. Univ., Moscow, 1968) [in Russian].
6.
Geological Map of the Elbrus Volcanic Area on a Scale of
1 : 50000
(FGUGP “SevKavGeologiya”, 1996).
7. G. V. Abikh, “Explanation of Geological Section of the
Northern Slope of the Caucasus Range from Elbrus to
Beshtau,” in
Caucasian Calendar for 1853
(Tiflis, 1852),
pp. 440–471.
8. E. E. Milanovskii and N. V. Koronovskii,
Orogenic Vol
canism and Tectonics of the Alpine Belt of Eurasia
(Nedra, Moscow, 1973) [in Russian].
9. A. M. Borsuk,
Mesozoic and Cenozoic Magmatic Forma
tions of the Greater Caucasus
(Nauka, Moscow, 1979)
[in Russian].
10. I. V. Chernyshev, V. A. Lebedev, S. N. Bubnov, et al.,
“Isotopic Geochronology of Quaternary Volcanic
Eruptions in the Greater Caucasus,” Geokhimiya,
No. 11, 1–16 (2002) [Geochem. Int.
40
, 1042–1055
(2002)].
11. V. A. Lebedev, I. V. Chernyshev, S. N. Bubnov, et al.,
“Chronology of Magmatic Activity of the Elbrus Vol
cano (Greater Caucasus): Evidence from K–Ar Iso
tope Dating of Lavas,” Dokl. Akad. Nauk
405
(3),
389–394 (2005) [Dokl. Earth Sci.
405
, 1321–1326
(2005)].
12. O. A. Bogatikov, I. V. Melekestsev, A. G. Gurbanov, et
al., “Radiocarbon Dating of Holocene Eruptions of the
Elbrus Volcano in the Northern Caucasus, Russia,”
Dokl. Akad. Nauk
363
(2), 219–221 (1998) [Dokl.
Earth Sci.
363
, 1093–1095 (1998)].
13. O. A. Bogatikov, I. V. Melekestsev, A. G. Gurbanov, et
al., “The Elbrus Caldera in the Northern Caucasus,”
Dokl. Akad. Nauk
363
(4), 515–517 (1998) [Dokl.
Earth Sci.
363
, 1202–1204 (1998)].
14. O. A. Bogatikov, I. V. Melekestsev, A. G. Gurbanov, et
al., “Catastrophic Pleistocene and Holocene Activity
of the Elbrus Volcanic Center (North Caucasus, Rus
sia): Events and Chronology Based on
14
C, EPR, and
K–Ar Dating,” Vulkanol. Seismol., No. 2, 3–17
(2001).
15. O. A. Bogatikov, A. G. Gurbanov, D. G. Koshchug, et
al., “The EPR Dating of the RockForming Quartz
from Volcanic Rocks of the Elbrus Volcano, Northern
Caucasus, Russia” Dokl. Akad. Nauk
385
(1), 92–96
(2002) [Dokl. Earth Sci.
385
, 570–573 (2002)].
16. O. A. Bogatikov, M. Ch. Zalikhanov, B. S. Karamurzov,
et al.,
Natural Processes on the Territory of Kabardino
Balkaria
(IGEM RAN, Moscow, 2004) [in Russian].
17. M. V. Abdulov, “On Geological Nature of the Elbrus
Gravity Anomaly,” Izv. Akad. Nauk SSSR, Ser. Geol.,
No. 9, 67–74 (1962).
18. N. V. Koronovskii and L. I. Demina, “Collisional Late
Cenozoic Volcanism of the Great Caucasus,” in
Pro
ceedings of 3rd AllRussia Symposium on Volcanology
and Paleovolcanology. Vol. 1. Volcanism and Geodynamics,
Ulan Ude, Russia, 2006
(Buryatsk. Nauchn. Ts. SO RAN,
UlanUde, 2006), pp. 219–223 [in Russian].
19. S. V. Vyatkin, Extended Abstract of Candidate’s Disserta
tion in Geology and Mineralogy (Mosk. Gos. Univ., Mos
cow, 2007).
20. I. V. Chernyshev, V. A. Lebedev, and M. M. Arakelyants,
“K–Ar Dating of Quaternary Volcanics: Methodology
and Interpretation of Results,” Petrologiya
14
(1), 69–89
(2006) [Petrology
14
, 62–80 (2006)].
21. I. V. Chernyshev, A. V. Chugaev, and K. N. Shatagin,
“HighPrecision Pb Isotope Analysis by Multicollector
ICPMassSpectrometry Using
205
Tl/
203
Tl Normaliza
tion: Optimization and Calibration of the Method for the
Studies of Pb Isotope Variations,” Geokhimiya, No. 11,
1155–1168 (2007) [Geochem. Int.
45
, 1065–1076 (2007)].
22. R. H. Steiger and E. Jager, “Subcommission on Geochro
nology: Convention on the Use of Decay Constants in
Geo and Cosmochronology,” Earth Planet. Sci. Lett.
36
,
359–362 (1977).
23. V. G. Molyavko, I. M. Ostafiichuk, and N. V. Koro
novskii, “Evolution, Chemistry, and Genesis of the Elbrus
Volcanics,” Izv. Akad. Nauk SSSR, Ser. Geol., No. 6, 31–
46 (1980).
24 . V. M. Gaz eev, A . A. Noso va, L . V. S azo nova, et a l., “ Petr o
genetic Interpretation of Phenocryst Association in the
Pleistocene–Holocene Volcanics of Elbrus (North Cau
casus),” Vulkanol. Seismol., No. 2, 24–45 (2004).
25. J. A. Pearce, N. B. W. Harris, and A. G. Tindle, “Trace
Element Discrimination Diagrams for the Tectonic Inter
pretation of Granitic Rocks,” J. Petrol.
25
, 956–983
(1984).
26. J. B. Whalen, K. L. Currie, and B. W. Chapell, “AType
Granites: Geochemical Characteristics, Discrimination
and Petrogenesis,” Contrib. Mineral. Petrol.
95
, 407–419
(1987).
27. B. R. Frost, C. G. Barnes, W. J. Collins, et al., “A
Geochemical Classification for Granitic Rocks,” J. Petrol
42
, 2033–2048 (2001).
28. V. A. Lebedev, S. N. Bubnov, I. V. Chernyshev, et al.,
“Geochronology and Genesis of Subalkaline Basaltic
Lava Rivers at the Dzhavakheti Highland, Lesser Cauca
sus: K–Ar and Sr–Nd Isotopic Data,” Geokhimiya, No.
3, 243–258 (2007) [Geochem. Int.
45
, 211–225 (2007)].
29. V. A. Lebedev, I. V. Chernyshev, A. V. Chugaev, et al.,
“K–Ar Age and Sr–Nd Characteristics of Subalkali
Basalts in the Central Georgian Neovolcanic Area
(Greater Caucasus)," Dokl. Akad. Nauk
408
(4), 517–522
[Dokl. Earth Sci.
408
, 657–661 (2006)].
30. V. A. Lebedev, S. N. Bubnov, I. V. Chernyshev, et al.,
“Geochronology and Genesis of the Young (Pliocene)
Granitoids of the Greater Caucasus: Dzhimara Mul
tiphase Massif of the Kazbek Neovolcanic Area,”
Geokhimiya, No. 6, 582–602 (2009) [Geochem. Int.
47
,
550–567 (2009)].

GEOCHEMISTRY
INTERNATIONAL
Vol. 48
No. 1
2010
GEOCHRONOLOGY OF ERUPTIONS AND PARENTAL MAGMA SOURCES 67
31. B. G. Polyak, I. L. Kamenskii, E. M. Prasolov, et al.,
“Helium Isotopes in Gases of the Northern Caucasus:
Implications for Heat and Mass Influx from the Mantle,”
Geokhimiya, No. 4, 383–397 (1998) [Geochem. Int.
36
,
329–342 (1998)].
32. A. W. Hofmann, “Mantle Geochemistry: the Message
from Oceanic Volcanism,” Nature
385
(16), 219–229
(1997).
33. Yu. A. Kostitsyn and A. A. Kremenetskii, “Age of Final
Magmatic Stage of the Eldjurtu Granite: Rb–Sr Isochron
Dating of Apatites,” Geokhimiya, No. 7, 925–931 (1995).
34. A. Z. Zhuravlev and E. V. Negrei, “Synchronous Forma
tion of the Eldzhurta Granite and OreBearing Metaso
matites of Tyrnyauz (North Caucasus) Based on Rb–Sr
Data,” Dokl. Akad. Nauk SSSR
332
(4), 483–487 (1993).
35. I. R. Pohl, J. S. Hess, B. Kober, et al., “Origin and Petro
genesis of Miocene Trachyrhyolites (Atypes) from the
Northern Part of the Greater Caucasus,” in
Magmatism of
Rifts and Foldbelts
(Nauka, Moscow, 1993), pp. 108–125
[in Russian].
36. S. B. Jakobsen and G. J. Wasserburg, “Sm–Nd Isotopic
Evolution of Chondrites and Achondrites,” Earth Planet.
Sci. Lett.
67,
137–150 (1984).
37. A. Zindler and S. Hart, “Chemical Geodynamics,” Annu.
Rev. Earth Planet. Sci., No. 14, 493–571 (1986).
38. P. D. Kempton, R. S. Harmon, C. J. Hawkesworth, et al.,
“Petrology and Geochemistry of Lower Crustal Granu
lites from the Geronimo Volcanic Field, Southeastern Ari
zona,” Geochim. Comochim. Acta
54
(12), 3401–3426
(1990).
39. K. H. Wedepohl,
Handbook of Geochemistry. Vol. II/4. 82
(Lead)
(Springer–Verlag, Berlin–Heidelberg–New York,
1974).
40. I. V. Chernyshev, M. M. Arakelyants, V. A. Lebedev, et al.,
“K–Ar Isotope Systematics and Age of Lavas from the
Quaternary Kazbek Volcanic Province, Greater Cauca
sus,” Dokl. Akad. Nauk
367
(6), 810–814 (1999) [Dokl.
Earth Sci.
367
, 862–866 (1999)].