GeolCarp_Vol57_No5_379

GEOLOGICA CARPATHICA, OCTOBER 2006, 57, 5, 379—396

www.geologicacarpathica.sk

Upper mantle xenoliths from the Pliocene Kozákov volcano

(NE Bohemia): P-T-f

and geochemical constraints

PATRIK KONEČNÝ

, JAROMÍR ULRYCH

, PAVEL SCHOVÁNEK

, MONIKA HURAIOVÁ

and ZDENĚK ŘANDA

Geological Survey of the Slovak Republic, Mlynská dolina 1, 817 04 Bratislava, Slovak Republic; konecnyp@gssr.sk

Institute of Geology, Academy of Sciences of the Czech Republic, Rozvojová 269, 165 02 Praha 6, Czech Republic

Czech Geological Survey, Klárov 131/3, 118 21 Praha 1, Czech Republic

Comenius University, Department of Mineralogy and Petrology, Mlynská dolina, 842 15 Bratislava, Slovak Republic

Nuclear Physics Institute, Academy of Sciences of the Czech Republic, 250 68 Řež, Czech Republic

(Manuscript received November 9, 2004; accepted in revised form October 6, 2005)

Abstract: Upper mantle xenoliths are abundant in the basanites of the Pliocene Kozákov volcano. The studied samples
of spinel lherzolites come from the depth of about 50—75 km. Their mineral assemblage preserved subsolidus tempera-
tures of 1165—1052

ºC from the time of xenolith entrapment. Oxygen fugacity varies from +0.14 to +0.93 log unit

relative to fayalite-magnetite-quartz (FMQ) buffer. Major bulk-rock oxides and variations in mineral chemistry indicate
a continual depletion trend mainly associated with extraction of basaltic melt from the mantle. Mineralogical features and
the absence of highly oxidized lherzolites suggest a negligible degree of modal metasomatic overprint. On the contrary,
the LREE upward patterns and U-shaped REE patterns of clinopyroxenes, as well as of the bulk lherzolite compositions
are indicators of cryptic metasomatic event(s) in the upper mantle. The U-shape REE patterns corroborates to enrichment
mechanisms in the mantle by reactive porous flow and chromatographic fractionation. A possible cryptic metasomatic
event(s) could have occurred in pre-Cenozoic times, probably during the Variscan orogeny.

Key words: Bohemian Massif, Cenozoic volcanism, ultramafic xenoliths, spinel lherzolite, harzburgite, peridotite.

Introduction

The young continental intraplate volcanism (Upper Creta-
ceous to Quaternary) of the Bohemian Massif forms an in-
tegral part of the Central European Volcanic Province.
The origin and distribution of the volcanism are con-
trolled by WSW—ENE-oriented tectonic systems related to
the Ohře/Eger Rift and by the WNW—ESE Labe Tectono-
volcanic Zone (Fig. 1). On the basis of radiometric ages
and geochemical signatures, Ulrych & Pivec (1997) and
Ulrych et al. (1999) defined two series of intraplate alka-
line volcanisms in the Bohemian Massif: (i) a pre-rift
unimodal ultramafic (79—49 Ma) and (ii) a rift-related bi-
modal volcanism (42—0.26 Ma).

Even though mantle xenoliths of the peridotite—pyroxen-

ite series are abundant in young volcanics of the Bohemian
Massif (more than 100 occurrences have been reported, see
Ulrych & Adamovič 2004), they have still not been system-
atically studied (cf. Jakeš & Vokurka 1987).

Ultramafic xenoliths from basanitic lava flows of the

Kozákov volcano, situated in the Labe Tectono-volcanic
Zone (Fig. 1), NE Bohemia, were studied by Fediuk
(1971), Vokurka & Povondra (1983), Vokurka & Kober
(1993), Medaris et al. (1997, 1999) and Christensen et al.
(2001). The lava flows are widely known by the presence
of frequent and large ultramafic xenoliths. Farský (1876)
already published the very first chemical analyses of oliv-
ine, orthopyroxene, clinopyroxene and spinel of the
Kozákov lherzolite xenoliths, thereby preceding the gen-
eral interest in mantle-derived materials by about one cen-

tury. The recent studies on lherzolite xenoliths from the
Kozákov volcano documented two textural types of xeno-
liths: (i) medium-grained equigranular and (ii) coarse-
grained protogranular, crystallized at 975—1090 ºC and
1.20—1.86 GPa (Medaris et al. 1997). Studies on ultramafic
xenoliths from other regions of the Bohemian Massif were
presented by Fediuk & Fediuková (1989), Frýda &
Vokurka (1995) and Ulrych et al. (2000).

The upper mantle beneath the Central European Volca-

nic Province reveals (i) vertical and lateral heterogeneities
(cf. Babuška & Plomerová 1988) and (ii) metasomatic in-
fluence (Lloyd 1987; Wilson & Downes 1991; Wedepohl
et al. 1994; Wilson et al. 1994; Downes 2001). These
variations commonly correspond to the enrichment/deple-
tion of upper mantle lherzolite in incompatible elements
as reported, by Wedepohl (1987), Downes (2001), Downes
et al. (1992), and Wedepohl et al. (1994) from the Central
European and the Western European Volcanic Provinces.

Based on textural and geochemical evidence, Lloyd &

Bailey (1975) postulated a metasomatic transition be-
tween spinel lherzolite and alkali clinopyroxenite
(clinopyroxene + titaniferous phlogopite ± titanian magne-
tite, apatite, titanite and rare corroded olivine) for the xe-
nolith suites from the rift valley volcanics of the West
Eifel and Uganda. It has been suggested that metasoma-
tism is either a necessary precursor to magmatism (Lloyd
& Bailey 1975; Wass & Rogers 1980; Witt-Eickschen &
Kramm 1998) or a consequence of alkali basaltic
magmatism (Wilshire et al. 1980; Menzies et al. 1985).
The likely connection between the origin of alkali

380

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

clinopyroxenite xenoliths and intrusive alkali complexes
was emphasized by Upton (1967), Lloyd & Bailey (1975,
1994) and Erickson et al. (1985). These authors consid-
ered them manifestations of the same igneous processes.
Compared with the original lherzolite mantle, pyroxenite
xenoliths indicate enrichment in incompatible and large
ion lithophile elements, and the host lavas are thought to
be indicative of the rheomorphic metasomatized mantle
(Lloyd & Bailey 1975).

Seismological data indicate a total lithosphere thick-

ness of 80—140 km beneath the Bohemian Massif, of
which about 70—80 km are possibly formed by highly
anisotropic olivine-rich mantle lithosphere (Babuška &
Plomerová 1988; Plomerová et al. 1998).

Variations in the upper mantle composition beneath the

Bohemian Massif can be deduced from geochemical data
on peridotite mantle xenoliths in the above mentioned
Cenozoic volcanic rocks. Peridotite mantle xenoliths in
the Bohemian Massif are free of K-, OH-, F-bearing phases
from the supposedly metasomatized mantle. Using Sr iso-
topic data, Frýda & Vokurka (1995) mentioned carbonate
metasomatism in a lherzolite xenolith from a basanite flow
of the central part of the České středohoří Mts. Phlogopite
and Mg-hastingsite were described by Kramer & Seifert
(2000) in dunite xenoliths from the Cenozoic basaltic
rocks of the Elbe Zone, Saxony. Lherzolite—harzburgite
xenoliths in basanite dykes of the České středohoří Mts
and basanite intrusion at Plesý Hill in the Lužické hory

Mts (Ulrych & Adamovič 2004) are rarely accompanied
by independent kaersutitic hornblende xenocrysts.

Clinopyroxenite forms rare layers in lherzolite xenoliths

and scarce clinopyroxenite discrete nodules in lavas of the
Kozákov volcano (Vokurka & Povondra 1983; Medaris et
al. 1997). Clinopyroxenite xenoliths of similar composition
were described by Mihaljevič (1993) from the basanite flow
in the central part of the České středohoří Mts, Ulrych et al.
(2000) from the Osečná Complex in the lateral block of the
Ohře Rift in N Bohemia, and by Ulrych et al. (2005) from
the Loučná—Oberwiesenthal Volcanic Centre in a similar
structural position in the Krušné hory Mts, (Fig. 1).

The aim of the paper is to contribute to the estimation of

P-T-f

conditions and to decipher the geochemical char-

acter of the upper mantle beneath the Kozákov volcano
based on the study of the major, minor and trace element
chemistry of xenoliths of different lithologies, textural
types and their mineral phases.

Geological setting

The Kozákov volcano (743 m a.s.l., see geological

sketch in Figs. 1 and 2) is situated in NE Bohemia on an
arm of the Lužice (Lusatian) fault bordering the Labe
Tectono-volcanic Zone, a prominent Cenozoic structure
of the Bohemian Massif. Lava flows of the Kozákov vol-
cano have been exposed in many quarries. Three lava

Fig. 1. A tectonic sketch of the Bohemian Massif and the main Cenozoic tectonic structures, volcanic and sedimentary deposits (modi-
fied after Christensen et al. 2001).

381

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

flows of the Early Pliocene ages (3.95 Ma for the upper-
most, 4.14 Ma for the lowermost flows in the Smrčí—Proseč
area (cf. Šibrava & Havlíček 1980) were recognized (cf.
Fediuk 1968). Wilson et al. (1994) presented the age of
4.25 Ma for the lowermost flow from Slap quarry. Never-
theless, ages of 6.39 Ma and 6.60 Ma for the lowermost
flows were published by Bellon & Kopecký (1977) and
Šibrava & Havlíček (1980), respectively. The lowermost
lava flow can be divided into an upper, thicker part with
thin columnar entablature, and a basal false colonnade
with thick columns (Fediuk in Ulrych et al. Eds. 1991).
Radiometric ages of the lava flow correspond to its geo-
logical position (overlying old terrace sediments of the
Jizera River) and stratigraphical position (palynological
evidence for Middle Miocene age – Konzalová 1973).

Methods of study

Although ultramafic mantle xenoliths are present in all

lava flows of the Kozákov volcano, they can be collected
from the lowermost flow of the Smrčí—Proseč—Slap area
only. Previously collected samples from the locality of
Pelechov were used for a comparison, see Fig. 2. Quarries
in other areas of the Kozákov volcano with the uppermost
lava flows have been closed many years ago. The xeno-
liths were studied macroscopically in the quarries in much
detail, and microscopically in thin sections.

Quantitative (magnetic and heavy liquid) separation of

mineral fractions of olivine, orthopyroxene, clinopyroxene
and spinel was applied to the study of two ellipsoidal
lherzolite xenoliths from two localities: Smrčí – 1.727 kg,
and Slap – 4.175 kg. Separation was performed on 1100 g
of crushed original rock from Smrčí and 1500 g from Slap,
using the most frequent size fractions of 0.125—0.250 mm.
Final concentrates were tested for purity (98—99 %) using
an optical method.

The composition of mineral phases of ultramafic xeno-

liths was determined using a CamScan 4—Link-eXL elec-
tron microprobe (Analyst R. Rybka, Czech Geological
Survey, Praha) with an acceleration voltage of 15 kV,
beam current of 20 nA, beam size about 3—5 mm, counting
time 20 s for peaks and 5 s for background, both natural
and synthetic standards and the ZAF reduction program.

Major element concentrations in the host rock, ultramafic

xenoliths and their mineral fractions were determined using
wet chemical methods (Analyst P. Povondra, Charles Uni-
versity, Praha). Analyses of the international reference
whole-rock standards (GM, TB, BN) and duplicate analyses
of the samples suggest total errors of ± 5 % (1 s.d.).

Trace element concentrations were determined by in-

strumental neutron activation analysis (INAA). Powdered
samples 50—90 mg in weight were sealed in polyethyl-
ene foils and irradiated for 4 hours in a nuclear reactor
LWR-15 of the Institute of Nuclear Research at Řež. The
neutron flux rate was 8.5 10

—2

· s

—1

. The irradiated

Fig. 2. A sketch of the geological setting of the Kozákov volcano lava flows (modified after Kopecký 1968).

382

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

samples were measured after three days and four months.
The gamma-ray spectrometric system, equipped with a
HPGe-detector, was operated with the following parameters:
efficiency 22 % and resolution FWHM 1.8 keV for photons
1332.5 keV of

Co. The counting time was 1.5 hours in the

first run and 24 hours in the second one. The precision
(1 s.d.) is better than ± 10 % for the REE and ± 5 to 10 % for
other trace elements. The accuracy of the data was checked
against international rock standards. It is better than 10 %
for the REE and 5 % for the other trace elements.

A quadrupole-based inductively coupled plasma mass

spectrometry (ICP-MS) (VG Elemental PQ3, UK) was used
for determination of REE and other trace element con-
tents (Analyst L. Strnad, Charles University, Praha). Pa-
rameter Values: Rf power (W) – 1420, gas flows (l/min
Ar 5.0) – 13.5, 1.3, 0.72, acquisition time of 3.50 s, nebu-
lizer of Meinhard type, acquisition mode – peak jump,
points per peak – 3, dwell time – 10.24, replicates – 3,
detector mode – pulse, internal standards –

Sc,

115

In,

185

Re. The precision (1 ) for incompatible elements was

better than ± 5 %. Accuracy was checked using above
mentioned international reference rock standards. Detec-
tion limits are in µg/l (ppb).

Characteristics of the xenoliths

Ultramafic xenoliths constitute about 99 % of all xeno-

liths in basaltic flows of the Kozákov volcano (Fediuk
1968). Harzburgites and lherzolites (about 96 %) prevail
over dunites (3 %) and ortho- and clinopyroxenites (1 %).
Modal compositions calculated from bulk-rock composi-
tions and mineral compositions plot the ultramafic xeno-
liths from the Kozákov volcano to the lherzolite field, less
frequently to the harzburgite, dunite and wehrlite fields
(Fig. 3).

On the basis of field observations, the proportion of el-

lipsoidal xenoliths 1 to 70 cm in size was estimated at
2 vol. % of the total rock volume (Fediuk 1968). The total
volume of xenolithic material is likely approaching
10 vol. % if discrete nodules and xenocrysts are included.
About 50 % of all xenoliths were 1—2 cm in size, xeno-
liths larger than 5 cm represent 10 % (Fediuk 1968). In
many cases, however, the original size of the xenoliths
was originally larger than given above as the xenoliths
are angular fragments originated during magma transport
and/or solidification of the host magma.

The contact of the xenoliths with the host rock is mostly

sharp. The reaction rim at the contact of the xenolith with
the host rock is 0.3 to 0.8 mm wide only. A broader
clouded amphibolized zone was observed only around
orthopyroxene at contact with the host rock. Rare elonga-
tion of olivines and pyroxenes in xenoliths (?) was ob-
served, partly highlighted by spinel arrangement to
elongated clusters. Microscopic study of the xenoliths
confirmed two distinct types of textures: a dominant me-
dium- to coarse-grained protogranular and minor medium-
grained equigranular type and rare porphyroclastic types
(Table 1) (sensu Mercier & Nicolas 1975).

Mineral chemistry

Mineral chemistry was studied using four methods: wet

analyses of pure mineral fractions (99—100 %) used for the
trace element study, ICP-MS, INAA (Table 2), and micro-
probe analyses in thin sections (Tables 3 and 4).

Olivine

reveals generally high forsterite contents

(Fo

89.4—91.8

) compatible with the composition of upper

mantle olivines. The variation of Fo content in olivines
within singular samples is small. A comparison of the
chemical composition of olivine between samples indicates
certain correlations: the lowest Fo values were encountered
in lherzolite S4x with porphyroclastic texture ( ~ Fo

89.2

low to intermediate values were found in protogranular
lherzolites (Fo

89.4—90.6

), and the highest values in

harzburgites (Fo

91.2—91.8

). No significant differences were

found either between cores and rims or between small and
large crystals.

Orthopyroxene and clinopyroxene

are both highly

magnesian, showing narrow ranges of Mg-number [(Mg/
(Mg + Fe)*100] of 88.4—93.7 and 89.4—91.6, respec-
tively. Orthopyroxenes from harzburgites yield higher
values (91.0—93.7), whereas lherzolites yield lower values
(88.3—91.4). In contrast, Mg-numbers of clinopyroxenes in
harzburgites (89.4—90.6) and lherzolites (89.5—91.5)

Fig. 3. Classification diagram for peridotites (Streckeisen 1979). Modal
abundances calculated by least square fitting of measured mineral
composition to the bulk-rock analyses. Where the respective minerals
were not analysed, average data from literature on compositions of
phases from ultramafic xenoliths were used. SiO

, TiO

, Al

, Cr

FeO

, MnO, MgO, CaO, and Na

O contents were involved in multilin-

ear regression, r

was better than 0.98 for all samples.

383

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

Table 1: List of samples of ultramafic xenoliths and their host rocks from the Kozákov volcano.

Table 2: Trace element analyses (in ppm) of minerals from protogranular lherzolite and host basanite from Slap, Kozákov volcano. Modal
composition according to separated mineral phases: xenolith 67 (ICP-MS) – olivine 58 %, clinopyroxene 12 %, orthopyroxene 25 %,
spinel 5 % and xenolith USo (INAA) – olivine 56 %, clinopyroxene 15 %, orthopyroxene 24 %, spinel 5 %. SM and SL (INAA) – bulk-
rock analyses, n.d. – not determined.

384

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

Table 3:

epresentative analyses and number of ions per formula unit calculated to 4 oxygens (olivine – ol) and 3 cations (spinel – sp)

of ultramafic xenoliths from the Kozákov volcano.

– wet analyses, all other microprobe analyses,

n.a. – not analysed.

385

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

Table 4:

Representative

analyses

orthopyroxene

(opx)

and

clinopyroxen

(cpx)

and

number

ions

per

formula

unit

calculated

gens

ultramafic

xenoliths

from

the

Kozá-

kov

volcano.

–

wet

analyses,

all

other

microprobe

analyses,

n.a.

–

not

anal

ysed.

USo

v S2

v S1

x S4

4x 6

USo

era

x op

54.

53.

54.

55.

53.

52.

70 54.

10 56.

61 53.

72 54.

29 56.

52.

07 50.

66 53.

44 52.

92 51.

27 50.

51.

53.

15.

32.

27.

31.

33.

00 33.

04 33.

37 32.

45 33.

74 34.

17.

26 18.

93 16.

99 17.

43 16.

26 17.

16.

17.

16.

19.

18.

19.

21.

22.

21.

19.

20.

.a. n

.a.

a. n

.a. n

.a.

.76

ota

99.

100.

99.

69 100.

98.

99.

01 99.

92 98.

30 99.

14 99.

99.

59 99.

72 99.

83 99.

67 99.

05 98.

99.

98 100.

31 100.

100.

895

872

927

936

918

881

836

893

952

887

892

942

891

848

941

929

876

857

863

879

907

920

000

003

002

003

001

002

001

000

007

000

007

003

002

158

189

118

106

188

172

244

145

102

193

160

098

224

220

160

152

249

228

233

200

239

206

226

182

467

458

145

154

115

149

171

161

166

170

096

121

143

145

096

097

109

102

085

083

002

004

000

006

004

003

004

005

003

001

003

010

009

005

003

006

003

002

713

701

456

483

649

719

714

723

715

699

753

762

935

030

920

947

887

936

918

927

911

887

005

054

046

028

057

076

085

093

030

041

030

016

748

729

774

781

834

855

859

847

741

808

012

003

000

002

007

005

001

090

049

000

073

061

003

001

000

003

008

000

.000

000

013

020

000

014

015

027

016

011

019

015

011

034

041

022

021

026

027

022

034

030

025

000

ati

027

026

014

011

979

024

026

025

990

010

023

004

025

049

969

984

980

009

010

998

992

993

053

050

028

022

000

047

052

049

000

020

045

007

049

097

000

017

019

000

849

854

813

810

814

854

863

857

817

836

850

826

889

920

840

848

852

874

873

864

855

858

100

88.

90.

75.

76.

91.

93.

92.

90.

91.

90.

89.

86.

90.

89.

90.

91.

42.

38.

42.

41.

45.

42.

45.

88.

87.

74.

75.

89.

88.

89.

6 87.

7 89.

5 89.

4 90.

0 90.

52.

5 54.

8 50.

1 50.

6 48.

8 49.

48.

49.

52.

49.

11.

23.

386

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

overlap. The lowest Mg-number for orthopyroxene was
recorded in protogranular herzolite USo, in agreement
with its fertile character reflected in the incompatible ele-
ment contents. The positive trend of calcium content ex-
pressed as the number of Ca (p.f.u.) in orthopyroxene and
clinopyroxene in Fig. 4 points to the coexistence rela-
tions of both pyroxenes. Clinopyroxenes from lherzolites
USo and S4x are poor in number of Ca (p.f.u.) probably
due to their incomplete re-equilibration. Better positive
correlation was found in Al

wt. % contents between

the two pyroxenes (Fig. 5), supporting pyroxene coexist-
ence.

Spinels

are typically chromium-rich, with Cr

con-

tents varying between 21.8 and 34.1 wt. %. The ratio of

Cr/(Cr + Al + Fe

3 +

)*100 in protogranular lherzolites (S1x,

P1x) is lower (22.3—27.9) than in harzburgites (37.1—40.4).
Spinels from porphyroclastic xenolith S4x are very hetero-
geneous (23.9—37.3), suggesting a non-equilibrium origin
of the sample. Unusual spinel from progranular lherzolite
USo with 17.9 wt. % Cr

and extremely high Al

con-

tent (68.0) belongs to the Al-spinel group.

Variations in spinel composition are presented in Fig. 6.

Protogranular and equigranular xenoliths recognized by
Medaris et al. (1999), Christensen et al. (2001), and dis-
played as fields in Fig. 6, show a narrow scatter of data
compared to the studied xenoliths. Fine-grained
harzburgite P4x plots close to the equigranular lherzolite
field. Spinels of other lherzolites plot near the
protogranular lherzolite field. Spinel in porphyroclastic
lherzolite S4x with high Cr-number approaches spinels
from the equigranular field in its chemical composition,
and the other spinels approach those from the
protogranular field. The two types of spinel from
protogranular lherzolite S4x connected with a dashed line
show large differences in the Cr/(Cr + Al) ratio.

Bulk-rock analyses of major elements

Chemical composition of the studied lherzolite xeno-

liths (see Table 5) shows systematic variations if the ele-
ments are plotted against MgO, which is a sensitive
indicator of the degree of depletion. Basalt-compatible
oxides SiO

, CaO, Al

, TiO

correlate negatively with

MgO, as was observed in many xenoliths worldwide (e.g.
Downes 2001 from the Central European Volcanic Prov-
ince; Luhr & Aranda-Gómez 1997; Xu et al. 1998;

Fig. 4. Correlation between the number of Ca (p.f.u.) in clinopy-
roxene and orthopyroxene. A positive correlation was achieved
except for samples USo and S4x.

Fig. 5. Correlation between Al

contents (wt. %) in coexisting ortho-

and clinopyroxenes. Temperature data are listed in Table 5. The tem-
perature of 1090

ºC delimits the upper protogranular and the underly-

ing equigranular mantle layer defined by Medaris et al. (1999).

Fig. 6. Compositional variations in spinel in terms of Cr/(Cr + Al)
vs. Mg/(Mg + Fe

2 +

) ratios. Shaded areas denote xenoliths from the

Kozákov volcano after Medaris et al. (1999). The dashed line con-
nects two types of spinel with remarkably different Cr/(Cr + Al) ra-
tios in porphyroclastic lherzolite S4x.

387

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

Parkinson & Pearce 1988). Linear correlations of Al

NiO and CaO vs. MgO contents for the studied xenoliths
clearly visible in Fig. 7 (data listed in Table 5) can be re-
lated to the depletion by multiple melt extraction from the
mantle. Straight-line correlations between SiO

, TiO

and

O vs. MgO contents are not very significant. The

source material for melting is supposed to be the primitive
average mantle. None of the studied xenoliths is close to
the average primitive mantle composition, suggesting that
all samples represent residua depleted after partial melt-
ing. The most strongly depleted samples with the highest
MgO content (olivine discrete xenoliths – PAL and USx)
plot on the opposite side of the depletion trend having the
lowest Al

, TiO

, SiO

and CaO concentrations. Nickel

content increases with the degree of depletion except for
sample PAL, indicating supplemental processes modify-
ing Ni content. Samples from Chuchelna (C) and Smrčí (S)
are remarkably enriched in CaO, with contents even ex-
ceeding those of the primitive mantle, while SiO

is rela-

tively low. Such high concentrations cannot be explained
by variations in mineralogy, but rather correspond to the
presence of calcite microveinlets also associated with al-
teration in subsolidus conditions.

Distribution of REE and other trace elements in

the mafic xenoliths and their clinopyroxenes

Trace element data on four separated mineral phases

from protogranular lherzolite 67 and USo are given in
Table 2. Bulk-rock contents were calculated using modal
abundances of four constituent mineral phases and their
trace element contents. The bulk-rock Ni and Co con-
tents (2114 and 108 ppm) are comparable with those in

lherzolites from continental rift zones and those from
oceanic within-plate settings summarized by Siena &
Coltorti (1993). Scandium accumulates preferentially in
clinopyroxene and olivine, whereas Ni and Co accumu-
late in olivine only.

The C1-normalized bulk-rock REE pattern is governed

by the contribution of clinopyroxenes (McDonough et al.
1992a; Xu et al. 1998). The trace element analyses of the
clinopyroxene and the host rock (Table 2) show a clear co-
herency of C1-normalized REE pattern, but with different
abundances (Fig. 8).

The C1-normalized REE pattern of clinopyroxene from

the Kozákov lherzolite USo presented in Fig. 8 shows rela-
tive LREE enrichment with (La

/Lu

)

cpx

= 9 .6. Both the

bulk rock and clinopyroxene in protogranular lherzolite
USo display a flat REE pattern from MREE to HREE, but
with apparent LREE enrichment. Similar REE patterns of
lherzolites from the Western and Central European Volca-
nic Provinces (Downes 2001), Wangqing, China (Xu et al.
1998), Zabargad, Red Sea (Piccardo et al. 1993), Mt
Lessini and Antarctica (Siena & Coltorti 1993), Balaton
Highland Volcanic Field, Hungary (Downes et al. 1992)
document LREE enrichment in metasomatized lherzolites.
The observed LREE-rich pattern is comparable with REE
pattern in clinopyroxenes from equigranular xenoliths de-
pleted in basaltic compositions (CaO < 2 wt. %), in which
orthopyroxene prevails over clinopyroxene (Downes et al.
1992; Xu et al. 1998; Downes 2001). Protogranular
lherzolite USo is LREE-enriched, with CaO content of
~2.5 wt. %, an average Mg-number of ~0.89, and a higher
Al/Si ratio, which is conformable with lherzolites/
harzburgites enrichment in incompatible elements.

C1-normalized REE pattern of protogranular lherzolite

67 is different. It has a U-shaped pattern similar to the

Table 5: Chemical analyses (wt. %) of peridotite xenoliths and host basanite from the Kozákov volcano: Mg-number – Mg/(Mg + Fe

)*100,

n.d. – not determined, n.a. – not analysed.

388

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

protogranular or in transitional between protogranular and
equigranular xenoliths found in Eifel, Germany (Stosh et
al. 1986), Massif Central, France (Downes & Dupuy 1987;
Lenoir et al. 2000), Wangqing, China (Xu et al. 1998).

Upward LREE inflection and the U-shaped C1-normal-

ized REE patterns, in the case of Kozákov lherzolites, sug-
gests metasomatic processes modifying the original mantle
composition.

Fig. 7. Major oxides and minor element contents of bulk xenoliths
plotted against MgO. Normalizing values of primitive mantle (PM)
from Sun (1982). Grey fields correspond to compositions of spinel
peridotite xenoliths from Mexico associated with Cenozoic rift zone
located on a thin continental crust (Luhr & Aranda-Gómez 1997)
and empty fields bounded by dashed line are peridotites from Vo-
gelsberg, Rhine Graben, Germany (Witt-Eickschen 1993).

389

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

Fig. 8. C1-normalized REE patterns of lherzolites, clinopyroxenes,
host basanite from the Kozákov volcano and of primitive mantle
(McDonough et al. 1992b). Normalizing values of C1 chondrite
from Taylor & McLennan (1985). Two types of pattern are clearly
distinguishable: (1) LREE enriched, protogranular lherzolite USo
and equigranular lherzolite 94KZSM4 from Smrčí (unpublished
data of E. Jelínek) and (2) U-shape REE pattern in protogranular
lherzolite 67. Dashed lines contour 10—30 % batch partial melting
residuum of primitive mantle, partition coefficients taken from Ion-
ov et al. (2002).

P-T-f

conditions

Spinel lherzolite assemblages do not allow reliable pres-

sure calculations to be made. The only available barometer,
based on Ca partitioning between olivine and clinopyroxene
(Köhler & Brey 1990), yields a wide pressure range, almost
independent on equilibration temperatures, with the P-T
path being non-conformable with the geothermal gradient
(Medaris et al. 1999; Konečný et al. 1999). The main reason

Table 6: Two-pyroxene temperatures calculated after Köhler &
Brey (1990), arranged according to increasing temperature, con-
fronted with Mg-number of olivine.

for this is the different diffusion rate of Ca in both miner-
als, much higher in olivine than in clinopyroxene. During
fast cooling of mantle xenoliths after a rapid ascent of the
host magma, the Ca distribution in both minerals is inho-
mogeneous. Pressure calculations are then spurious. Errors
may also be introduced by analytical methods. Measure-
ment of low Ca concentrations in olivine calls for special
care with microprobe calibration or the use of secondary ion
mass spectrometry (SIMS) analyses. Instead of “absolute”
pressure, a reasonable pressure range might be deduced
from the position of spinel/garnet and spinel/plagioclase
phase transition. The upper pressure limit is constrained by
the Moho discontinuity. Spinel is stabilized by an increas-
ing Cr content (O’Neill 1981; Webb &Wood 1986). Calcu-
lated pressures of the spinel/garnet phase boundary for
spinels with Cr-number of 0.2—0.4 using the method of
O’Neill (1981) range from 2.4 to 3.0 GPa. The present thick-
ness of the crust beneath the Kozákov volcano is about
32 km (Čermák et al. 1991). Consequently, using an aver-
age density of 3.3 g · cm

—3

, the xenoliths are collected from

depths of ~ 32 km down to 74 km (the most aluminium-rich
spinel) to 92 km (the most chromium-rich spinel).

The two-pyroxene thermometer (Brey & Köhler 1990)

was used for the calculation of equilibrium temperatures. A
constant pressure of 1.5 GPa was assumed. All calculated
temperatures from 10 samples (Table 6) lie in a very narrow
range, from 1052 to 1165 ºC. The lowest temperatures
were recorded in pyroxenites F1, F2. The porphyroclastic
lherzolite xenolith S4x reveals the highest temperature,
thereby confirming its complex origin. Temperatures do not
seem to depend strictly on the type of the xenolith, whether
compared with harzburgite or lherzolite.

Oxygen fugacity is a parameter characterizing the redox

state of the xenoliths and consequently also of the upper
mantle. Oxygen fugacity conditions depend on the con-
tent of ferric iron in spinel, being in equilibrium with oxy-
gen exchange between spinel, olivine, orthopyroxene and
clinopyroxene. One of the most accepted versions of oxy-
gen fugacity calculation (Ballhaus et al. 1991) was used.
The pressure of 1.5 GPa and the temperature 1100 ºC (av-
erage temperature of our two-pyroxene thermometer calcu-
lations) were taken as the mean mantle conditions beneath
the Kozákov area. The range of oxygen fugacity calcu-
lated for four samples clusters above the fayalite-magne-

390

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

tite-quartz (FMQ) buffer (from + 0.14 to + 0.93 FMQ, Fig. 9)
with the exception of harzburgite P4x, which plots one
and a half units below FMQ. All samples in Fig. 9 with the
exception of P4x lie in the field or very close to the field
for slightly metasomatized lherzolites. No sample is of
“primitive” character or highly oxidized by metasomatic re-
actions. The chemical composition of harzburgite P4x is
anomalous as it plots outside the fields in Fig. 8, indicat-
ing that it is in a state of oxygen disequilibrium. The re-
dox state of porphyroclastic lherzolite S4x indicates
equilibrium for the spinel with Cr/(Cr + Al) of 0.26 and dis-
equilibrium for the Cr-rich spinel, with Cr/(Cr + Al) of 0.54.

Discussion

Depletion in xenoliths

The occurrence of upper mantle xenoliths of different

lithologies and textural types in a single volcanic vent is a
common feature of many volcanic localities worldwide. A
depletion process in the mantle is responsible for varia-
tions in modal composition, major, minor and trace ele-
ment contents and different textural types of ultramafic
xenoliths. The modal composition of rocks derived from
mantle peridotite gradually changes, passing from fertile
lherzolites to harzburgites and dunites. The presence of
different upper mantle xenoliths in basanite flows of the
Kozákov volcano is linked to an inhomogeneity of the
upper mantle caused by the extraction of variable volumes
of basaltic melt from the primary upper-mantle material.

Fig. 9. Deviations of oxygen fugacity from the FMQ buffer plot-
ted against Cr-numbers (Cr/Cr + Fe) in spinel. The fields are com-
piled from many data sources presented by Ballhaus et al. (1991).

However, its possible primary compositional inhomoge-
neity cannot be completely excluded. The chemical
composition of the upper mantle xenoliths and the com-
position of the constituent minerals is proportional to
the degree of partial melting in the Kozákov volcano re-
gion: the most abundant are lherzolites, which predomi-
nate over harzburgites, together comprising about 96 % of
all ultramafic xenoliths; dunites, websterites and pyrox-
enites are rare (Fediuk 1971). The refractory character is
linked with increase in Mg-number and decrease in CaO
and Al

contents. CaO/Al

ratios above 1 correspond

to lherzolites, and those below 1 to more refractory
harzburgites. The lowest TiO

, Al

, CaO, Cr

con-

tents and the highest MgO contents were recorded in the
most residual and refractory dunite. All samples, except
for dunites, have MgO contents confined to a small range,
being quite far from the primitive mantle. This may indi-
cate either a similar degree of melt extraction and a certain
degree of homogeneity in the upper mantle, or similar
depths of xenolith entrapment.

Almost all samples have a Ca/Al ratios less than 1.1,

characteristic for chondrites (Table 5), indicating a partial
melting process, which increases the Ca/Al ratio whereas
Al decreases (Palme & Nickel 1985). The lowest Ca/Al ra-
tio indicates the most depleted dunite PAL. If we exclude
samples Smrčí (S) and Chuchelna (C), which have ex-
tremely high CaO content (Fig. 7) and consequently do
not follow the general fractionation trend, then only
lherzolite No. 69 has the Ca/Al ratio above 1.1, suggesting
an enrichment in clinopyroxene probably by modal meta-
somatism in the mantle (Stein & Katz 1989).

Variations in modal compositions of the xenoliths are ac-

companied by changes in chemical compositions of the
minerals. Chromium content in spinel is a significant indi-
cator of the depletion process (Bonatti & Michael 1989;
Siena et al. 1991). The depletion trend and metasomatic pro-
cess can be recognized by negatively correlated Al

con-

tents in coexisting orthopyroxenes and Cr/(Cr + Al)*100
ratios in spinels, see Fig. 10. The chemical composition of
xenoliths, except for porphyroclastic lherzolite S4x (cf.
Fig. 10), follow the depletion trend originally defined for
conditions of continental passive margins (Bonatti &
Michael 1989). The most intensive depletion is exhibited
by harzburgite P4x, lying outside the continental lherzolite
field (Fig. 10) on an extension toward a higher Cr/(Cr + Al)
ratio and a lower Al

content in spinel and orthopyroxene,

respectively. The most fertile lherzolite is represented by
protogranular lherzolite USo.

The compositions of spinel and olivine during progres-

sive melting of lherzolite approach those of Cr-rich spinel
and Fo-rich olivine. No trend in terms of Cr/(Cr + Al + Fe

3 +

)

ratios in spinel and Fo-contents in olivine is evident (Fig. 11),
but it may rather follow compositional differences of xeno-
liths, cf. harzburgite P4x vs. lherzolites (all other samples).

Metasomatic event(s) in the mantle

The process of mantle depletion resulting in geochemical

and/or mineralogical changes in upper mantle composition

391

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

is well documented. Metasomatic event(s) caused by reac-
tion between infiltrating agents (melts and/or supercritical
fluids) and the mantle rocks are often superimposed on de-
pleted mantle. Previous studies, based on the origin, com-
position and relation to geotectonic setting, have identified
three types of metasomatizing agents: subduction-related
fluids (Andersen et al. 1984; Ayers 1998; Zanetti et al.
1999), carbonate melts (Thibault et al. 1992; Ionov et al.
1993; Green & Vallace 1998; Neumann et al. 2002) and
mafic silicate melt (Bodinier et al. 1990; Siena et al. 1991;
Ionov et al. 1998). The mantle beneath the Kozákov area

Fig. 10. Compositions of orthopyroxene and spinel of ultramafic xe-
noliths from the Kozákov volcano in a diagram of Al

contents in

orthopyroxene vs. Cr/(Cr + Al)*100 ratio in spinel, with trends of de-
pletion and metasomatism. Heavy line fields – oceanic peridotites,
continental peridotites and peridotites of subduction margins after
Bonatti & Michael et al. (1989), dashed line fields – peridotites
from southern Slovakia after Konečný et al. (1995), related to Qua-
ternary alkaline volcanism associated with lithosphere thinning and
mantle upwelling during Pliocene—Pleistocene.

Fig. 11. Cr/(Cr + Al + Fe

3 +

) ratios in spinel vs. Fo contents in oliv-

ine. Grey fields represent equigranular and protogranular xenoliths
from the Kozákov volcano after Medaris et al. (1999).

was located underneath a Cenozoic tectono-volcanic zone,
thus we rule out subduction-related fluids. Interaction with
carbonatite melt may lead to the formation of wehrlites
(Yaxley et al. 1991) or, under different P-T conditions,
carbonatite melt may quench in veins or pockets (Ionov et
al. 1998) or primary carbonates can crystallize. The Ti/Eu
ratio is insensitive to different degrees of partial melting
and preserves primitive mantle values of ~ 7740 (Yaxley et
al. 1998). Invasion of carbonatite melt substantially de-
creases that ratio in peridotite. Mineralogical evidence of
carbonatite metasomatism is absent in the Kozákov peridot-
ites, which is consistent with their high Ti/Eu ratio of 6700
for protogranular lherzolite USo and 11500 for protogranular
lherzolite 67. The passage of silicate magma through the
mantle peridotites is typically accompanied by formation of
LREE-enriched amphiboles and/or clinopyroxenes. Silicate
magma mostly moves through a system of veins. The inten-
sity of LREE, Sr and Zr enrichment decreases towards the
edge of the veins and in the wallrock peridotite it may di-
minish within some centimeters (Witt-Eickschen et al.
1998). Clinopyroxenite and/or amphibole veins were not
observed in the Kozákov xenoliths and modal metasoma-
tism is almost entirely absent. Nevertheless, clinopyroxene
and bulk-rock REE patterns (USo protogranular lherzolite)
show enrichment in LREE exactly as would be expected for
modal metasomatism by silicate melts. LREE enrichment is
largely linked to clinopyroxene, but not with newly formed
clinopyroxene. Most probably, a pervasive metasomatizing
agent penetrated the volume of the peridotite and modified
the composition of the original clinopyroxenes. Microscale
exchange reactions took place by diffusion over a longer
time interval. Such LREE enrichment in clinopyroxene
solely by diffusion and/or exchange reaction has been
termed “cryptic metasomatism” (Dawson 1984; Picardo et
al. 1993; Wiechert et al. 1997).

The U-shape of REE pattern of protogranular xenolith

67 is very similar to those modelled by chromatographic
fractionation and reactive porous flow during passage of
silicate melts through the peridotite matrix (Navon &
Stolper 1987; Takazawa et al. 1992; Nielson & Wilshire
1993; Godard et al. 1995). The melt equilibrates with the
peridotite matrix by a diffusion process. Melt percolating
for longer distances is selectively enriched in highly in-
compatible elements due to chromatographic reactions
with host peridotite. Changes in the composition of the
silicate melt with distance from the melt source, for ex-
ample a vein system, result in gradation from a steady
LREE enriched shape in clinopyroxenes to a U-shape far
from the melt source. Either numerical calculations in-
volving 1-D porous flow (constant porosity) or percolative
fractionation model (decreasing porosity with distance)
can satisfactorily match the REE patterns at various dis-
tances from the melt source (Ionov et al. 2002).

The occurrence of two types of spinel (porphyroclastic

lherzolite S4x) with significantly different chromium con-
tent indicates a bulk metasomatic overprint. The position of
high-chromian spinel outside of the general depletion trend
and given fields of continental or oceanic peridotites is
compatible with a secondary metasomatic origin (cf. Fig. 9).

392

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

The observations of cryptic and possible instance of lo-

cal modal metasomatism, suggests that at least some por-
tions of the mantle beneath Kozákov underwent one or
more metasomatic event(s) before fragmentation and en-
trainment of xenoliths by alkaline magma.

Evolution and conditions

Medaris et al. (1999) and Christensen et al. (2001) pre-

sented evidence for a layered structure of the upper mantle
beneath the Kozákov volcano area. The structure consists of
three horizontal layers, with protogranular lherzolites sand-
wiched between equigranular lherzolites. The protogranular
lherzolite layer spans a depth of 32—70 km. A layered struc-
ture of the lithospheric mantle was constrained by compari-
son of shear-wave splitting of xenoliths and teleseismic
observations, which revealed horizontal foliation in the up-
per equigranular layer, and vertical foliation in the middle
protogranular and the lower equigranular mantle layer
(Medaris et al. 1999; Christensen et al. 2001). The middle
protogranular layer represents a part of a slab of garnet peri-
dotites tectonically emplaced into the upper mantle during
Variscan convergent motions (Medaris et al. 1997). Garnet
in the protogranular layer became metastable and was re-
placed by stable spinel-pyroxene symplectite. Thermal con-
ditions in the lithospheric mantle were modified by
underplating of the 5 km thick basaltic magma reservoir at
30 Ma age. The shallow upper mantle was heated to higher
temperatures and subsequent cooling was faster than in
deeper parts.

Although the depth of xenolith extraction cannot be

calculated reliably for spinel lherzolite assemblages, we
can force the temperature data to fit on a model geotherm
of 5 Myr (Medaris et al. 1999). The pressures estimated by
such a method give a depth range of 60—80 km. This range
is far below the Moho at the present time in N Bohemia
(about 32 km – Čermák et al. 1991) and is near the
spinel-garnet phase transition boundary (74—92 km).

For the middle protogranular layer, use of the Brey &

Köhler (1990) two-pyroxene thermometer, is limited to
the temperature range of 850—1090 ºC (Medaris et al.
1999). The temperatures of the Kozákov xenoliths ar-
ranged in ascending order (1052—1165 ºC) indicate that
most samples refer to the lower part of the protogranular
slab and a few high-temperature samples correspond to
the underlying equigranular layer. Samples correspond-
ing to the upper equigranular layer are absent. The high-
est temperatures are recorded in samples containing
Al-rich coexisting pyroxenes (Fig. 4).

The Mg-numbers of olivine (90.6—91.8) tend to in-

crease with temperature and pressure, defining a trend of
progressive depletion with depth. Medaris et al. (1999)
have reported an exactly opposite correlation, suggest-
ing that no straightforward relation exists between depth
and degree of depletion.

In regions with strong metasomatic overprint, the oxy-

gen fugacity is raised above the FMQ buffer up to +3 log
units (Ballhaus et al. 1991; Amundsen & Neumann 1992).
The oxidation state of the upper mantle beneath the

Kozákov volcano was slightly above the FMQ buffer
(from + 0.14 to + 0.93 log unit), which compares well with
the oxygen fugacity for the subcontinental lithospheric
mantle ( + 0.24 ± 0.5) from the classical localities of the
world (Massif Central, Kilbourne Hole, San Carlos, Cen-
tral Asia – Wood & Virgo 1989; Bryndzia & Wood
1990). The oxidation state of the mantle beneath the
Kozákov volcano implies a restricted role for modal meta-
somatism. On the contrary, the presence of Cr-rich spinels
(Cr/(Cr + Al) > 0.5) is most probably the result of increased
temperature during metasomatism, but the rare occurrence
of these spinels decreases the significance of this process.
The cryptic metasomatism has a negligible effect on the
calculation of oxygen fugacity.

Two processes considered to promote Cr/(Cr + Al) ratio in

spinel are depletion and metasomatism. The harzburgite
P4x is relatively reduced but has high Cr/(Cr + Al) ratio in
spinel (Fig. 8) and additionally is the most depleted xeno-
lith. Increased chromium in spinel may be the result of more
intensive depletion than in the majority of the non-
metasomatized peridotites (Fig. 8).

Protogranular lherzolite USo shows evidence of ex-

posure to bulk cryptic metasomatism. According to the
equilibrium temperature of this xenolith (1086 ºC), it
originates from the bottom part of the protogranular
mantle layer, which was not thermally affected by Ceno-
zoic rifting (Medaris et al. 1999). Cryptic metasomatism re-
quires a longer time to change the composition of the
clinopyroxenes, compared to flow infiltration of melt
through a vein system reacting directly with the sur-
rounding mantle. Diffusion model calculations of reac-
tion between metasomatizing melt and clinopyroxene,
modifying the inherited LREE-depleted to an LREE-en-
riched pattern, imply that complete re-homogenization
of a 1 mm crystal takes up to 16 kyr (Beccaluva et al.
2001). By contrast, amphibole in peridotite adjacent to a
metasomatizing melt may attain the observed diffusion pro-
file in a period of not longer than 10—15 years (Witt-
Eickschen et al. 1998). The effectiveness of liquid-solid
diffusion reactions may imply the occurrence of metasomatic
reactions a short time before the xenolith entrainment in the
host basalt (Witt-Eickschen et al. 1998; Wulf-Pedersen et al.
1999). A much slower process(es) for cryptic metasomatism
may be expected, as a much smaller volume is involved in
reaction with the mantle peridotite. It has been suggested
that the metasomatizing magmas are to be derived from the
deep asthenospheric mantle (Schiano & Clocchiatti 1994;
Vaselli et al. 1995; Witt-Eickschen et al. 1998).

Protogranular lherzolite 67, according to temperature

1095 ºC, is located at the boundary between the
protogranular and underlying equigranular mantle layer
(Medaris et al. 1999). The U-shaped REE pattern of
protogranular lherzolite 67 is intrepreted as consequence of
cryptic metasomatism of highly fractionated silicate melts
with the host peridotite. Melt enriched in incompatible ele-
ments can be evolved over relatively short distances
(Zanetti et al. 1996), but Nielson & Wilshire (1993) argue
for long percolation paths necessary for effectiveness of the
chromatographic fractionation. A single event modelled by

393

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

reactive porous flow can explain the majority of both types
of REE patterns in clinopyroxenes (LREE enriched and U-
shaped) in the Spitsbergen mantle. The process operated be-
neath the volcanic area on a scale of several kilometers
involving a large portion of the lithospheric mantle (Ionov
et al. 2002). A large scale modal metasomatic event in the
mantle beneath the Quaternary West Eifel Volcanic Field,
Germany, led to the formation of secondary amphibole,
clinopyroxene and phlogopite during the Variscan orogeny
(Shaw et al. 2005). A younger event, related to Quater-
nary volcanism, is characterized by amphibole-phlogopite-
clinopyroxene veins (Shaw et al. 2005).

The similar depth of protogrnanular xenoliths USo and

67 cannot be an argument for a large scale pervasive meta-
somatism in the lithospheric mantle under Kozákov. Both
xenoliths come from middle protogranular mantle layer, de-
fined by Medaris et al. (1999), intensively affected during
the Variscan orogeny. The astenospheric metasomatizing
melts could have been generated during the Variscan orog-
eny, when initial collisional movements were followed by
post-thickening extension, which resulted in uplift of the
protogranular mantle layer to the recent position. Either the
protograular mantle layer was metasomatized prior to final
emplacement or synchroneously with uplift, alternatively a
certain short time after uplift.

Arguments against the Neogene metasomatizing event

are: (1) gross mantle structure related most likely to the
Variscan tectonics (Christensen et al. 2001), (2) negligible
Neogene reheating of the protogranular mantle layer to pro-
mote diffusion metasomatic reactions, (3) a relatively short
period of mantle relaxation after uderplating and hence
possible time for infiltration of deep astenospheric melts,
(4) absence of clinopyroxene-amphibole ± phlogopite veins,
hosted by peridotite, associated with passage of silicate
melts during Quaternary volcanism.

Conclusions

The ultramafic xenoliths from the Kozákov volcano rep-

resent a deeper part of a layered uppermost mantle as de-
scribed by Medaris et al. (1999) and Christensen et al.
(2001). Specifically the xenoliths correspond to the lower
half of the protogranular layer and to the underlying
equigranular layer, comprising a depth range of 50—75 km
and a temperature interval of 1165—1052 ºC. This deeper
mantle shares features of subcontinental mantle, deduced
from mineral compositions and oxygen fugacities clus-
tered slightly above the FMQ buffer (from + 0.14 to + 0.93
FMQ). The bulk-rock major element chemistry points to a
similar degree of depletion in the xenoliths, indicating a
coherent evolution of the Kozákov upper mantle. The de-
gree of depletion is sufficiently advanced and ubiquitous
for no single sample to approach the composition of the
primitive mantle. Subordinate dunites and wehrlites are
associated with mantle inhomogeneity.

The presence of two spinel types with low and high Cr

content in only one xenolith and absence of typical meta-
somatic minerals like amphiboles, clinopyroxenes or phlo-

gopites suggests negligible involvement of bulk metasoma-
tism. The LREE enriched pattern in clinopyroxenes from
protogranular lherzolites USo and the U-shaped REE pat-
tern in clinopyroxenes from protogranular lherzolites 67
reveal a modification of LREE, which cannot be ex-
plained by melt extraction but requires involvement of
cryptic metasomatic process(es). The exact age of the
metasomatic event(s) is unknown, but may have occurred
in pre-Neogene time, possibly in association with the late
Variscan development of the Bohemian Massif.

We conclude that, although there is very scant evidence

of modal metasomatism on a macro/micro scale, the REE
patterns in clinopyroxene and the bulk rock compositions
reflect cryptic metasomatic changes at least in some deeper
parts of the subcontinental mantle beneath Kozákov.

Acknowledgments:

Financial support for this research was

provided by Project A3013403 of the Grant Agency of the
Academy of Sciences of the Czech Republic, the Scien-
tific Programme CEZ: Z3-013-912 of the Institute of Geol-
ogy, AS CR, and subsidiary also from the VEGA Grant
No. 1/1029/04. The authors gratefully acknowledge L.G.
Medaris, Wisconsin, H.G. Stosh, Köln am Rhein, and F.E.
Lloyd, Bristol for their stimulating comments, and E.
Jelínek for providing unpublished data on xenoliths from
Kozákov. We thank O. Vaselli, Florence, Cs. Szabo,
Budapest and J. Lexa, Bratislava for their comments.

References

Amundsen H.E.F. & Neumann E.R. 1992: Redox control during

mantle/melt interaction. Geochim. Cosmochim. Acta 56,
2405—2416.

Andersen T., O’Reily S.Y. & Griffin W.L. 1984: The trapped fluid

phase in upper mantle xenoliths from Victoria, Australia: im-
plications for mantle metasomatism. Contr. Mineral. Petrology
88, 72—85.

Ayers J. 1998: Trace elements modeling of aqueous fluid – peri-

dotite interaction in the mantle wedge of subduction zones.
Contr. Mineral. Petrology 132, 390—404.

Babuška V. & Plomerová J. 1988: Subcrustal continental lithos-

phere: a model of its thickness and anisotropic structure. Phys.
Earth Planet. Inter. 51, 130—132.

Ballhaus C., Berry R.F. & Green D.H. 1991: High pressure experi-

mental calibration of olivine-orthopyroxene-spinel oxygen
geobarometer: implications for the oxidation state of the upper
mantle. Contr. Mineral. Petrology 107, 27—40.

Beccaluva L., Bonadiman C., Coltorti M., Salvini L. & Siena F.

2001: Depletion events, nature of metasomatizing agent and
timing of enrichment processes in lithospheric mantle xeno-
liths from the Veneto Volcanic Province. J. Petrology 42, 1,
173—187.

Bellon H. & Kopecký L. 1977: Spectres d’ages radiométriques du

volcanisme du rift du Massif Bohémien. 5

eme

Réunion Ann.

Sci. Terre, Rennes, Soc. Géol. Fr. éd. 57.

Bodinier J.L., Vasseur G., Verniéres J., Dupuy C. & Fabriés J.

1990: Mechanisms of mantle metasomatism: geochemical evi-
dence from Lherz orogenic massif. J. Petrology 31, 597—628.

Bonatti E. & Michael J. 1989: Mantle peridotites from continental

rifts to ocean basins to subduction zones. Earth Planet. Sci.
Lett. 91, 297—311.

394

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

Brey G.P. & Köhler T. 1990: Geothermobarometry in four-phase

lherzolites II. New thermobarometers, practical assessment of
existing thermobarometers. J. Petrology 31, 1353—1378.

Bryndzia L.T. & Wood B.J. 1990: Oxygen thermobarometry of

abyssal spinel peridotites: the redox state and C-O-H volatile
composition of the Earth’s suboceanic upper mantle. Amer. J.
Sci. 290, 1093—1116.

Čermák V.M., Král M., Kubík J. & Šafanda J. 1991: Heat flow, re-

gional geophysics and lithosphere structure in Czechoslovakia
and adjacent part of Central Europe. In: Čermák V. & Rybach
L. (Eds.): Terrestrial heat flow and lithosphere structure.
Springer-Verlag, Berlin, 133—165.

Christensen N.I., Medaris I.G., Wang H.F. & Jelinek E. 2001:

Depth variation of seismic anisotropy and petrology in central
European lithosphere: A tectonothermal synthesis from spinel
lherzolite xenoliths. J. Geophys. Res. 106, 645—664.

Dawson J.B. 1984: Contrasting types of upper mantle metasoma-

tism. In: Kornprobst J. (Ed.): Kimberlites. II. The mantle and
crust-mantle relationships. Elsevier, Amsterdam, 289—294.

Downes H. 2001: Formation and modification of the shallow sub-

continental lithospheric mantle: a review of geochemical evi-
dence from ultramafic xenolith suites and tectonically
emplaced ultramafic massifs of Western and Central Europe. J.
Petrology 42, 233—250.

Downes H. & Dupuy C. 1987: Textural, isotopic and REE varia-

tions in spinel peridotite xenoliths, Massif Central, France.
Earth Planet. Sci. Lett. 82, 121—135.

Downes H., Embey-Isztin A. & Thirlwall M. 1992: Petrology and

geochemistry of spinel peridotite xenoliths from western
Pannonian Basin (Hungary): evidence for an association be-
tween enrichment and texture in the upper mantle. Contr. Min-
eral. Petrology 109, 340—354.

Erickson S.C., Fourie P.J. & De Jager D.H. 1985: A cumulate ori-

gin for the minerals in clinopyroxenites of the Phalaborwa
Complex. Trans. Geol. Soc. S. Africa 88, 207—214.

Farský F. 1876: Mineralogische Notizen I. Mineralien aus der

Kosakover Basaltkugeln. K.-kön. Geol. Reichsanst, (Wien)
205—208.

Fediuk F. 1968: Additional notes on the basaltic volcanics and their

ultrabasic nodules at Smrčí in the Železný Brod area. Fac. Sci.
Charles Univ., Praha, 1—12 (in Czech).

Fediuk F. 1971: Ultramafitites in the area of the Krkonoše—Jizerské

hory Mountains. Acta Univ. Carol., Geol. 319—343 (in Czech).

Fediuk F. & Fediuková E. 1989: Ultramafic nodules in basalts from

northern Moravia, Czechoslovakia. Sbor. Geol. Věd, Ř. Geol.
44, 4—49.

Frýda J. & Vokurka K. 1995: Evidence for carbonatite metasoma-

tism in the upper mantle beneath the Bohemian Massif. J.
Czech Geol. Soc. 43, A—9—10.

Godard M., Bodinier J.L. & Vasseur G. 1995: Effects of mineral-

ogical reactions on trace element redistributions in mantle
rocks during percolation process: a chromatographic ap-
proach. Earth Planet. Sci. Lett. 133, 449—461.

Green D.H. & Vallace M.E. 1998: Mantle metasomatism by ephem-

eral carbonatite melts. Nature 336, 459—462.

Ionov D.A. 1998: Trace element composition of mantle-derived

carbonates and coexisting phases in peridotite xenoliths from
alkali basalts. J. Petrology 39, 1931—1941.

Ionov D.A., Dupuy C., O’Reily S.Y., Kopylova M.G. & Genshaft

Yu.S. 1993: Carbonated peridotite xenoliths from Spitsbergen:
implications for trace element signature of mantle carbonate
metasomatism. Earth Planet. Sci. Lett. 119, 283—297.

Ionov D.A., Bodinier J.L., Mukasa S.B. & Zanetti A. 2002: Mecha-

nisms and sources of mantle metasomatism: major and trace ele-
ment compositions of peridotite xenoliths from Spitsbergen in the
context of numerical modelling. J. Petrology 43, 2219—2259.

Jakeš P. & Vokurka K. 1987: Central Europe. In: Nixon P.H. (Ed.):

Mantle xenoliths. Chichester, 149—154.

Köhler T.P. & Brey G.P. 1990: Calcium exchange between olivine

and clinopyroxene calibrated as geothermobarometer for natu-
ral peridotites from 2 to 60 kb with applications. Geochim.
Cosmochim. Acta 54, 2375—2388.

Konečný P., Konečný V., Lexa J. & Huraiová M. 1995: Mantle xe-

noliths in alkali basalts of southern Slovakia. Acta Vulcanol. 7,
2, 241—247.

Konečný P., Huraiová M. & Bielik M. 1999: P-T-X-fO

conditions

in upper mantle: evidence from lherzolitic xenoliths hosted by
Plio-Pleistocene alkali basalts. Geolines 9, 59—66.

Konzalová M. 1973: Neogene plant microsossils of river sediments

from underground of the Neoidic volcanites of the Železný
Brod area. Věst. Ústř. Úst. Geol. 48, 17—23 (in Czech).

Kopecký P. 1968: Geological map of the territory between Železný

Brod, Semily and Kozákov. Unpublished MSc Thesis, Charles
University, Prague 1—88 (in Czech).

Kramer W. & Seifert W. 2000: Mafische Xenolithe und Magmatite

im östlichen Saxothuringikum und westlichen Lugikum: Ein
Beitrag zum Krustenbau und zur regional Geologie. Z. Geol.
Wiss. 28, 133—156.

Lenoir X., Garrido C.J., Bodinier J.L. & Dautria J.M. 2000: Con-

trasting lithospheric mantle domains beneath the Massif Cen-
tral (France) revealed by geochemistry of peridotite xenoliths.
Earth Planet. Sci. Lett. 181, 359—375.

Le Maitre R.W. (Ed.) 2002: Igneous rocks. A classification and

glossary of terms. 2

Edition. Cambridge University Press,

Cambridge, 1—236.

Lloyd F.E. 1987: Characterization of mantle metasomatic fluids in

spinel lherzolites and alkali clinopyroxenites from the West Eifel
and South Uganda. In: Menzies M.A. & Hawkesworth C.J.
(Eds.): Mantle metasomatism. Academic Press, London, 91—120.

Lloyd F.E. & Bailey D.K. 1975: Light element metasomatism of the

continental mantle: the evidence and the consequences. Phys.
Chem. Earth 9, 389—416.

Lloyd F.E. & Bailey D.K. 1994: Complex mineral textures in

bebedourite: possible links with alkali clinopyroxenite xeno-
liths and kamafugite volcanism. In: Mayer O.H.A. &
Leonardos O.H. (Eds.): Proceed. 5

Intern. Conf. Vol. I.

Kimberlites. Related rocks and mantle xenoliths. CPRM Spec.
Publ., Rio de Janeiro, Brazil, 263—269.

Luhr J.F. & Aranda-Gómez J.J. 1997: Mexican peridotite xenoliths

and tectonic terranes: correlations among vent location, tex-
ture, temperature, pressure and oxygen fugacity. J. Petrology
38, 2, 1075—1112.

McDonough W.F., Stosch H.S. & Ware N.G. 1992a: Distribution of

titanium and the rare earth elements between peridotitic miner-
als. Contr. Mineral. Petrology 110, 321—328.

McDonough W.F., Sun S.S. Ringwood A.E. Jagoutz E. & Hof-

mann. A.W. 1992b: Potassium, rubidium, and cesium in the
Earth and Moon and the evolution of the mantle of the Earth.
Geochim. Cosmochim. Acta. 56, 1001—1012.

Medaris L.G. (Jr.), Fournelle J.H., Wang H.F. & Jelínek E. 1997:

Thermobarometry and reconstructed chemical compositions of
spinel-pyroxene symplectites: evidence for pre-existing garnet
in lherzolite xenoliths from Czech Neogene lavas. Russian
Geol. Geophys. 38, 277—286.

Medaris L.G. (Jr.), Wang H.F., Fournelle J.H., Zimmer J.H. &

Jelínek E. 1999: A cautionary tale of spinel peridotite
thermobarometry: an example from xenoliths of Kozákov vol-
cano, Czech Republic. Geolines 9, 92—95.

Menzies M.A., Kempton P.D. & Dugan M.A. 1985: Interaction of

continental lithosphere and asthenospheric melts below the
Geronimo volcanic field, Arizona, U.S.A. J. Petrology 26,
663—693.

395

UPPER MANTLE XENOLITHS (NE BOHEMIA): P-T-f

AND GEOCHEMICAL CONSTRAINS

Mercier J.C. & Nicolas A. 1975: Textures and fabrics of upper-

mantle peridotites as illustrated by xenoliths from basalts. J.
Petrology 16, 454—487.

Mihaljevič M. 1993: Geochemistry of olivine-free ultramafic nod-

ules from the České středohoří Mts. Ph.D. Thesis Charles Uni-
versity, Prague, 1—112 (in Czech).

Navon O. & Stolper E. 1987: Geochemical consequences of melt

percolation: the upper mantle as chromatographic column. J.
Geol. 95, 285—307.

Neumann E.R., Wulf-Pedersen E., Pearson N.J. & Spencer E.A.

2002: Mantle xenoliths from Tenerife (Canary Islands): evi-
dence for reactions between mantle peridotites and silicic
carbonatite melts including Ca metasomatism. J. Petrology 43,
825—857.

Nielson J.E. & Wilshire H.G. 1993: Magma transport and metaso-

matism in the mantle: a critical review of current geochemical
models. Amer. Mineralogist 78, 1117—1134.

O’Neill H.S.T. 1981: The transition between spinel lherzolite and

garnet lherzolite, and its use as a geobarometer. Contr. Min-
eral. Petrology 77, 185—194.

Palme H. & Nickel K.G. 1985: Ca/Al ratio and composition of

the Earth’s upper mantle. Geochim. Cosmochim. Acta 49,
2123—2132.

Parkinson I.J. & Pearce J.A. 1998: Peridotites from the Izu-Bonin-

Mariana forearc (ODP Leg 125): evidence for mantle melting
and melt-mantle interaction in a supra-subduction zone setting.
J. Petrology 39, 9, 1577—1618.

Piccardo G.B., Rampone E., Vannucci R., Shimizu N., Ottolini L. &

Bottazzi P. 1993: Mantle processes in the sub-continental litho-
sphere: the case study of the rifted sp-lherzolites from
Zabargad (Red Sea). Europ. J. Mineral. 5, 1039—1056.

Plomerová J., Babuška V. & Horálek J. 1998: Seismic anisotropy

and velocity variation in the mantle beneath the
Saxothuringicum-Moldanubicum contact in Central Europe.
Pure Appl. Geophys. 151, 365—394.

Schiano P. & Clocchiatti R. 1994: Worldwide occurrence of silica-

rich melts in sub-continental and sub-oceanic mantle minerals.
Nature 368, 621—623.

Schovánek P. 1977: Petrology of selected ultramafic rocks of the

Bohemian Masssif and chemical composition of their minerals.
Unpublished PhD Thesis, Charles University, Prague 1—88 (in
Czech).

Shaw S.J., Eyzaguirre J., Fryer B. & Gagnon J. 2005: Regional

variations in the mineralogy of metasomatic assemblages in
mantle xenoliths from the West Eifel Volcanic Field, Germany.
J. Petrology 5, 945—972.

Šibrava V. & Havlíček P. 1980: Radiometric age of Plio-Pleistocene

volcanic rocks of the Bohemian Massif. Věst. Ústř. Úst. Geol.
55, 129—139.

Siena F. & Coltorti M. 1993: Thermobarometric evolution and

metasomatic processes of upper mantle in different tectonic
settings: evidence from spinel peridotite xenoliths. Europ. J.
Mineral. 5, 1073—1090.

Siena F., Beccaluva L., Coltorti M., Marchesi S. & Morra V. 1991:

Ridge hot-spot evolution of the Atlantic lithospheric mantle:
evidence from Lanzarote peridotite xenoliths (Canary Islands).
J. Petrology, Spec. Lherzolites Issue, 271—290.

Stein M. & Katz A. 1989: The composition of subcontinental litho-

sphere beneath Israel: inferences from peridotitic xenoliths. Is-
rael J. Earth Sci. 38, 75—87.

Stosh H.G. & Lugmair G.W. 1986: Trace elements and Sr and

Nd isotope geochemistry of peridotite xenoliths from Eifel
(W Germany) and their bearing on the evolution of the sub-
continental lithosphere? Earth Planet. Sci. Lett. 80, 281—298.

Stosh H.G., Lugmair G.W. & Kovalenko V.I. 1986: Spinel peridot-

ite xenoliths from the Tariat Depression, Mongolia. II:

Geochemistry and Sr isotopic trace elements and Nd a and Sr
isotopic composition and their implications for the evolution
of the sub-continental lithosphere. Geochim. Cosmochim. Acta
50, 2601—2614.

Streckeisen A. 1979: Classification and nomenclature of volcanic

rocks, lamprophyres, carbonatites and melilitic rocks: recom-
mendations and suggestions of the IUGS subcommision on the
systematics of igneous rocks. Geology 7, 331—335.

Sun S.S. 1982. Chemical composition and origin of the Earth’s

primitive mantle. Geochim. Cosmochim. Acta 46, 179—192.

Takazawa E., Frey F.A., Shimizu N., Obata M. & Bodinier J.L.

1992: Geochemical evidence form melt migration and reaction
in the upper mantle. Nature 359, 55—58.

Taylor S.R. & McLennan F. 1985: The continental crust: its compo-

sition and evolution. Blackwell, Oxford, 1—379 (in Russian).

Thibault Y., Edgar A. & Lloyd F. 1992: Experimental investigation

of melts from a carbonated phlogopite lherzolite: implications
for metasomatism in continental lithospheric mantle. Amer.
Mineralogist 77, 784—794.

Ulrych J. & Pivec E. 1997: Age-related contrasting alkaline volca-

nic series in North Bohemia. Chemie Erde 57, 311—336.

Ulrych J., Kopecký L. & Kropáček V. 1991(Eds.): Guide to Post-

symposium excursion: Neoidic volcanism of the Bohemian
Massif. June 27 to 29, 1991, Charles University 1991, 1—58.

Ulrych J., Lloyd F.E., Balogh K., Hegner E., Langrová A., Lang

M., Novák J.K. & Řanda Z. 2005: Petrogenesis of alkali py-
roxenite and ijolite xenoliths from the Tertiary Loučná—Ober-
wiesenthal Volcanic Centre, Bohemian Massif in the light of
new mineralogical, geochemical and isotopic data. Neu. Jb.
Mineral., Abh. 182, 57—79.

Ulrych J., Pivec E., Lang M., Balogh K. & Kropáček V. 1999:

Cenozoic intraplate volcanic series in of the Bohemian Massif:
a review. Geolines 9, 123—129.

Ulrych J., Pivec E., Povondra P. & Rutšek J. 2000: Upper-mantle

xenoliths in melilitic rocks of the Osečná complex, North
Bohemia. J. Czech Geol. Soc. 45, 1—15.

Upton B.G.J. 1967: Alkaline pyroxenites. In: Wyllie P.J. (Ed.): Ul-

tramafic and related rocks. J.Wiley, New York, 281—288.

Vaselli O., Downes H., Thirlwall M., Dobosi G., Coradossi N.,

Seghedi I., Szakacs A. & Vannucci R. 1995: Ultramafic xe-
noliths in Plio-Pleistocene alkali basalts from Eastern
Transylvanian Basin: Depleted mantle enriched by vein meta-
somatism. J. Petrology 36, 1, 23—53.

Vokurka K. 1979: The study of elements equilibrium between min-

erals of ultramafic xenoliths of the Kozákov volcano. Unpub-
lished MSc Thesis, Charles University, Prague (in Czech).

Vokurka K. & Kober B. 1993: Heterogeneity of the Earth’s mantle

below the Bohemian Massif. In: Kukal Z. (Ed.): Proceedings
of the 1

International Conference on the Bohemian Massif,

Prague, Czechosl. Sept. 26—Oct. 3, 1988. Czech Geol. Surv.,
Prague, 327—330.

Vokurka K. & Povondra P. 1983: Geothermometry and

geobarometry of lherzolite nodules from Kozákov, NE
Bohemia, Czechoslovakia. Acta Univ. Carol., Geol. 261—272.

Wass S.Y. & Rogers N.W. 1980: Mantle metasomatism – precursor

to continental alkaline volcanism. Geochim. Cosmochim. Acta
44, 1811—1823.

Webb S.A.C. & Wood B.J. 1986: Spinel-pyroxene-garnet relation-

ships and their dependence on Cr-Al ratio. Contr. Mineral. Pe-
trology 92, 471—480.

Wedepohl K.H. 1987: Kontinentaler Intraplatten-Vulkanismus am

Beispiel der tertiären Basalte der Hessischen Senke. Fortschr.
Mineral. 65, 19—47.

Wedepohl K.H., Gohn E. & Hartmann G. 1994: Cenozoic alkali ba-

saltic magmas of western Germany and their products of dif-
ferentiation. Contr. Mineral. Petrology 115, 253—278.

396

KONEČNÝ, ULRYCH, SCHOVÁNEK, HURAIOVÁ and ŘANDA

Wiechert U., Ionov D.A. & Wedepohl K.H. 1997: Spinel peridotite

xenoliths from the Atsagin-Dush volcano, Dariganga lava pla-
teau, Mongolia: A record of partial melting and cryptic meta-
somatism in the upper mantle. Contr. Mineral. Petrology 126,
346—367.

Wilshire H.G., Nielson Pike J.E., Mayers C.E. & Schwarzmann E.C.

1980: Amphibole rich veins in lherzolite xenoliths, Dish Hill
and Deadman Lake, California. Amer. J. Sci. A280, 576—593.

Wilson M. & Downes H. 1991: Tertiary-Quaternary extension-re-

lated alkaline magmatism in western and central Europe. J. Pe-
trology 32, 811—850.

Wilson M., Rosenbaum J. & Ulrych J. 1994: Cenozoic magmatism

of the Ohře rift, Czech Republic: geochemical signatures and
mantle dynamics. Abst. Int. Volcanol. Congress, Ankara 1.

Witt-Eickschen G. 1993: Upper mantle xenoliths from alkali basalts

of the Vogelsberg, Germany: implications for mantle upweling
and metasomatism. Eur. J. Mineral. 5, 361—376.

Witt-Eickschen G. & Kramm U. 1998: Evidence for the multiple

stage evolution of the subcontinental lithospheric mantle be-
neath the Eifel (Germany) from pyroxenite and composite py-
roxenite/peridotite xenoliths. Contr. Mineral. Petrology 131,
258—272.

Witt-Eickschen G., Kaminsky W., Kramm U. & Harte B. 1998: The

nature of young vein metasomatism in the lithosphere of the
West Eifel (Germany): geochemical and isotopic constraints
from composite mantle xenoliths from the Meerfelder Maar. J.
Petrology 34, 155—185.

Wood B.J. & Virgo D. 1989: Upper mantle oxidation state: ferric

iron contents of lherzolite spinels by

Fe Mössbauer spectros-

copy and resultant oxygen fugacities. Geochim. Cosmochim.
Acta 53, 1277—1291.

Wulf-Pedersen E., Neumann E.R., Vanucci R., Bottazzi P. &

Ottolini L. 1999: Silicic melts produced by reaction between
peridotite and infiltrating basaltic melts: ion probe data on
glasses and minerals in veined xenoliths from La Palma, Ca-
nary Islands. Contr. Mineral. Petrology 137, 59—82.

Xu Y., Menzies M.A., Vroon P., Mercier J.C. & Lin C. 1998: Tex-

ture-temperature-geochemistry relationships in the upper
mantle as revealed from spinel peridotite xenoliths from
Wangquing, NE China. J. Petrology 39, 469—493.

Yaxley G.M., Crawford A.J. & Green D.H. 1991: Evidence for

carbonatite metasomatism in spinel peridotites from western
Victoria, Australia. Earth Planet. Sci. Lett. 197, 305—317.

Yaxley G.M., Green D.H. & Kamenetsky V. 1998: Carbonatite

metasomatism in the Southeastern Australian lithosphere. J.
Petrology 39, 11, 12, 1917—1930.

Zanetti A., Vannucci R., Oberti P. & Ottolini L. 1996: Infiltration

metasomatism at Lherz as monitored by systematic ion-micro-
probe investigations close to a hornblendite vein. Chem. Geol.
134, 113—133.

Zanetti A., Mazzucchelli M., Rivalenti G. & Vannucci R. 1999: The

Finero phlogopite-peridotite massif: an example of subduc-
tion-related metasomatism. Contr. Mineral. Petrology 134,
107—122.