BIROn - Birkbeck Institutional Research Online

Downes, Hilary and Balaganskaya, E. and Beard, Andy and Liferovich, R.

and Demaiffe, D. (2005) Petrogenetic processes in the ultramafic, alkaline
and carbonatitic magmatism in the Kola Alkaline Province: a review. Lithos
85 (1-4), pp. 48-75. ISSN 0024-4937.

Downloaded from:

Usage Guidelines:
Please refer to usage guidelines at or alternatively
contact [email protected].
 Petrogenetic processes in the ultramafic, alkaline and carbonatitic magmatism in the
Kola Alkaline Province: a review
Hilary Downes1, Elena Balaganskaya2, Andrew Beard1, Ruslan Liferovich2, Daniel Demaiffe3
1
School of Earth Sciences, Birkbeck, University of London, Malet Street, London, UK
2
Geological Institute, Kola Science Centre RAS, Fersman Street, 14, Apatity, Murmansk oblast,
184200 RUSSIA
3
Laboratoire de Geochimie Isotopique (CP 160/02), Université Libre de Bruxelles, Ave. F
Roosevelt, B-1050 Bruxelles, Belgium

[Revised version accepted for publication in Lithos – June 2004]

Abstract
Igneous rocks of the Devonian Kola Alkaline Carbonatite Province (KACP) in NW Russia
and eastern Finland can be classified into four groups: (a) primitive mantle-derived silica-
undersaturated silicate magmas; (b) evolved alkaline and nepheline syenites; (c) cumulate rocks; (d)
carbonatites and phoscorites, some of which may also be cumulates. There is no obvious age
difference between these various groups, so all of the magma-types were formed at the same time in
a relatively restricted area and must therefore be petrogenetically related. Both sodic and potassic
varieties of primitive silicate magmas are present. On major element variation diagrams, the
cumulate rocks plot as simple mixtures of their constituent minerals (olivine, clinopyroxene, calcite
etc). There are complete compositional trends between carbonatites, phoscorites and silicate
cumulates, which suggests that many carbonatites and phoscorites are also cumulates. CaO/Al2O3
ratios for ultramafic and mafic silicate rocks in dykes and pipes range up to 5, indicating a very
small degree of melting of a carbonated mantle at depth. Damkjernites appear to be transitional to
carbonatites. Trace element modelling indicates that all the mafic silicate magmas are related to
small degrees of melting of a metasomatised garnet peridotite source. Similarities of the REE
patterns and initial Sr and Nd isotope compositions for ultramafic alkaline silicate rocks and
carbonatites indicate that there is a strong relationship between the two magma-types. There is also
a strong petrogenetic link between carbonatites, kimberlites and alkaline ultramafic lamprophyres.
Fractional crystallisation of olivine, diopside, melilite and nepheline gave rise to the evolved
nepheline syenites, and formed the ultramafic cumulates. All magmas in the KACP appear to have
originated in a single event, possibly triggered by the arrival of hot material (mantle plume?)
beneath the Archaean/Proterozoic lithosphere of the northern Baltic Shield that had been recently
metasomatised. Melting of the carbonated garnet peridotite mantle formed a spectrum of magmas
including carbonatite, damkjernite, melilitite, melanephelinite and ultramafic lamprophyre. Pockets
of phlogopite metasomatised lithospheric mantle also melted to form potassic magmas including
kimberlite. Depth of melting, degree of melting and presence of metasomatic phases are probably
the major factors controlling the precise composition of the primary melts formed.

Keywords: Kola Alkaline Carbonatite Province; alkaline ultramafic magmatism, carbonatites;

nepheline syenites, kimberlites
1. Introduction
The Kola Alkaline Carbonatite Province (KACP) is located in the northern Baltic Shield,
within the Kola Peninsula (Russia) and parts of NE Finland (Fig.1). The province is well-known for
its ultramafic, alkaline and carbonatite magmatism (Kogarko, 1987; Woolley, 1989; Arzamastsev,
1994; Kogarko et al., 1995; Bulakh et al., in press). It consists of nineteen alkaline-ultramafic rock
complexes, most of which also contain carbonatites, together with two giant nepheline syenite
complexes (Lovozero and Khibiny) and numerous dykes and pipes of various ultramafic and
evolved alkaline rocks, kimberlites and carbonatites. Calculations indicate that the total volume of
magma that formed the KACP was at least 15000 km3 (Arzamastsev et al., 2001), but this does not
take into account the possibility of an eroded lava cover. Although each magmatic group within the
KACP has been extensively investigated and discussed individually, this paper aims to discuss all
of the magmatic rock groups (ultramafic alkaline rocks, kimberlites, carbonatites and nepheline
syenites) together and to suggest a relationship between the different types of magmatism. We
present a review of the key petrological and geochemical features of the different magmatic groups
and discuss the origin and evolution of the KACP, using recently published papers that contain high
quality geochemical data. An Appendix provides a glossary of unusual rock-types found in the
KACP.

2. Geology and petrography

The northern Baltic Shield is composed predominantly of tonalite-trondhjemite-granodiorite

complexes formed around 2.95-2.90 Ga ago, with subsidiary supracrustal Late Archaean and Early
Proterozoic formations (Balagansky et al., 1998). The crust of the region is approximately 40km
thick and the present-day lithosphere thickness is estimated at >200km (see review by Artemieva
(2003)). During Palaeozoic times, the area was part of Baltica, a continent consisting of present-day
Scandinavia, Spitsbergen, Russia and Ukraine. Paleogeographic reconstructions show that Baltica
was situated in the southern hemisphere during Devonian times and was moving northward
(Torsvik et al., 1996). Several authors have linked the KACP with either a 'North Atlantic Alkaline
Province' that includes complexes in Greenland and the Canadian Shield (Vartiainen and Woolley,
1974) or with the extensive Devonian rifting and magmatism on the East European Platform
(Kramm et al., 1993). The age of the KACP magmatism is much more restricted than that of the so-
called 'North Atlantic Alkaline Province' and is much closer to the ages of the East European
Platform Devonian magmatism. The existence of a ENE-WSW trending rift system on the Kola
Peninsula (sometimes called the Kontozero rift) was first suggested by Kukharenko (1967) from the
general alignment of several KACP massifs. This trend also encompasses the outcrops of Devonian
volcano-sedimentary units (Arzamastsev et al., 1998). However, the generally N-S to NE-SW
orientation of the KACP dyke swarms on the south coast of the Kola Peninsula argues for an
approximately E-W extensional stress field at the time of their emplacement. Kukharenko et al.
(1971) and Vartiainen and Paarma (1979) suggested that the presence of several carbonatite
complexes and dyke swarms on the south coast of Kola was related to NNW-SSE trending
'Kandalaksha deep fracture system', which may be associated with the Mezen rifts beneath
Arkhangelsk.
The lithospheric mantle beneath the Kola Peninsula is represented by mantle xenoliths such
as spinel lherzolites and spinel wehrlites (Arzamastsev and Dahlgren, 1994; Beard et al., in prep).
These lithologies show evidence of metasomatism, as they contain 1-3% amphibole and ca. 1%
phlogopite, together with rare fluorapatite and picroilmenite. Kimberlites in Arkhangelsk (eastern
end of the White Sea) contain spinel peridotite and garnet peridotite xenoliths (Sablukov et al.,
2000), with rarer phlogopite clinopyroxenites, again indicative of metasomatic enrichment. The
lower continental crust of the Kola and Arkhangelsk region is composed mainly of mafic garnet
granulites (Markwick and Downes, 2000; Kempton et al., 2001) that are at least 2.4 Ga old
(Downes et al., 2000). Metasomatic phlogopite of 2.0 Ga age is present in some of these granulite
xenoliths (Kelley and Wartho, 2000).
 The Devonian magmatism was preceeded by several earlier periods of alkaline magmatic
activity in the northern Baltic Shield (see review by Bulakh et al., in press). Nepheline syenite
massifs of late Archaean age are found in the eastern part of the Kola Peninsula at Sakhariok and
Kuljok. Three early Proterozoic (ca. 2.0 Ga) alkaline massifs with associated carbonatites are found
in North Karelia (Tiksheozero and Eletozero) and western Kola (Gremyakha-Vyrmes). Potassic
magmatism also occurred in Kola and Karelia at approximately 1.7-1.8 Ga, followed by
emplacement of lamproitic dykes in Karelia at approximately 1.2 Ga. Uvad'ev (1980) reported
similar ages on alkaline lamprophyre dykes in the Kandalaksha Gulf. Elsewhere on the Baltic
Shield (e.g. Fen in southern Norway, Alnö in eastern Sweden, and the Kuopio-Kaavi kimberlites of
eastern Finland), there is evidence for late Proterozoic alkaline ultramafic magmatism at around 600
Ma. There are reports of similar ages for dykes within the Kandalaksha region (e.g. Rukhlov, 1999).
Thus, the Kola-Karelia region was the site of sporadic alkaline and carbonatitic magmatism long
before the main Devonian KACP magmatic event.

2.1 Dyke swarms and pipes

Two major Devonian dyke swarms are exposed on the south coast of the Kola Peninsula.
One swarm is located in the north-west part of the Kandalaksha Gulf and the other occurs on the
Turiy Peninsula (Fig. 1). Together they comprise more than 1000 bodies. Detailed descriptions are
given by Ivanikov (1977) and Rukhlov (1999). From field relations, confirmed by the sparse age
determinations (Fig. 2), dyke emplacement appears to have occurred in two stages in Early and Late
Devonian times. The Early Devonian dyke magmatism comprises alnöites, aillikites, damkjernites
and carbonatites, together with monticellite kimberlites and ultramafic lamprophyres (Mahotkin et
al., 2003). The Late Devonian magmatism, particularly developed in the Turiy Peninsula, consists
of dykes and pipes of kimberlites, alkaline and melilite picrites, melilites, nephelinites and
carbonatites (Ivanikov et al., 1998). Some of the dykes and pipes contain crustal and mantle
xenoliths that yield insights into the nature of the lithosphere beneath the region (Kempton et al.,
2001; Beard et al., in prep).
Borodin et al (1976) presented geochemical analyses of the wide variety of rock-types that
form the dykes and concluded that the lamprophyre dykes are closely related to the alkaline-
ultramafic complexes. They also suggested that the silicate magmas that formed the dykes were not
closely related to kimberlites. Carbonatite and ultramafic lamprophyre dykes from Kandalaksha
were investigated by Beard et al. (1996) and Mahotkin et al. (2003), whilst Kalinkin et al. (1993)
and Beard et al. (1998) studied pipes of sparsely diamoniferous kimberlite and melilitite from the
Terskiy coast. The petrogenesis of the melilitite-carbonatite-nephelinite dyke series of the Turiy
coast was discussed by Ivanikov et al (1998). Sindern et al (2004) have recently studied a suite of
late-stage alkaline picrite, monchiquite, nephelinite and phonolite dykes from the Khibiny complex.
A monticellite kimberlite dyke near Kandalaksha was regarded by Beard et al. (1998) as showing
some similarities to ultramafic lamprophyres. The diamondiferous kimberlite pipes of the
Arkhangelsk region (Mahotkin et al., 2000; Beard et al., 2000) are also Devonian in age and
probably belong to the KACP magmatism, despite being situated much further east (Fig. 1).

2.2 Ultramafic-alkaline carbonatite complexes (UACC)

The phoscorite-bearing ultramafic alkaline carbonatite complexes (UACC) of the KACP
have been extensively reviewed by Bulakh et al (in press), Krasnova et al (in press) and Wall and
Zaitsev (in press). The UACC are multiphase central intrusions with circular or elliptical exposures
and concentrically zoned structures (Kukharenko et al., 1965; Kogarko et al., 1995). Their centres
contain a series of ultramafic cumulus rocks, whereas melilite-bearing rocks and strongly silica-
undersaturated feldspathoidal rocks (foidolites) occur nearer to the periphery. In some complexes a
multistage phoscorite-carbonatite series forms concentrically zoned stockworks. Dyke rocks are
mostly alkaline and nepheline syenites along with monchiquites, tinguaites and ultramafic
lamprophyres, which crosscut all other rock-types. The distribution of rock-types within the various
complexes is shown in Table 1. The only complex in which all the above-mentioned rocks are
found is Kovdor, which is the best exposed occurrence and also the one that has been studied in
most detail (Krasnova and Kopylova, 1988; Balaganskaya, 1994; Krasnova, 2001). Table 2 shows
details of the sequence of magmatic rocks within the Kovdor complex (Liferovich, 1998). Verhulst
et al (2000) presented a detailed study of the petrology and geochemistry of Kovdor. A review of
the phoscorite and carbonatite occurrences in Kovdor is given by Krasnova et al. (in press).
Several recent papers have dealt with the Kola UACC. The Kandaguba massif was
investigated by Pilipiuk et al (2001). Balaganskaya et al. (1999a), Brassinnes et al (2003 and in
press) and Karchevsky and Moutte (in press) produced new data on the Vuoriayarvi complex,
Vartianen (1980) and Lee et al (in press) have discussed the Sokli complex, Zaitsev et al (in press)
and Sitnikova (2004) have reported on the Sallanlatvi complex, and Afrikanda has been studied by
Chakhmouradian and Zaitsev (in press). The Turiy UACC (Fig. 1) was also the subject of a recent
major study by Dunworth and Bell (2001), while Karchevsky (in press) deal with the phoscorite-
carbonatite part of this complex. Lapin (1979), Bulakh et al. (1998) and Balaganskaya et al (2000)
studied the Seblyavr complex. Arzamastsev et al (2000a) gave details of gravity and density
modelling of the size and shape of the hidden portions of all the major UACC intrusions.
The ultramafic rock group includes cumulates composed of olivine and clinopyroxene, with
minor cumulus Ti-rich magnetite and perovskite and intercumulus Ti-garnet, amphibole,
phlogopite, apatite and calcite. Proportions of cumulus and intercumulus phases vary widely, and
cumulate textures have often been obliterated by replacement of clinopyroxene by late magmatic
mica and amphibole. Melilitic rocks are also highly unusual cumulate lithologies in which the
proportions of major minerals vary widely, including cumulates with >85% melilite (melilitolites).
The most abundant rock-type is turjaite, which consists of cumulus melilite, nepheline,
clinopyroxene and phlogopite, with accessory olivine, Ti-rich magnetite, Ti-garnet, amphibole,
perovskite and minor apatite and calcite. Foidolites form a series of cumulates with differing
proportions of cumulus clinopyroxene, nepheline/cancrinite and rarer melilite, forming melteigites,
ijolite-urtites and urtites. The rocks are highly heterogeneous in terms of textures, grain-size and
mineral composition. Interstitial calcite is normally in equilibrium with nepheline, indicating that it
is a primary igneous mineral.
Carbonatites occur in all periods of formation of the UACC, including the early magmatic
one in which phoscorites are absent, in a late magmatic phase where they are found intimately
related with phoscorites, and in a post-magmatic period (see Table 2). Early carbonatites form
dykes up to 150 m wide, cross-cutting the ultramafitolites, foidolites and fenites. In addition to
calcite or dolomite, they contain feldspathoids, biotite/phlogopite, clinopyroxene/amphibole and
titanite. In some cases they have both clear contacts with host silicate rocks but elsewhere they
gradually merge into cumulate clinopyroxenite or melteigite-ijolite (Kukharenko et al., 1965;
Ternovoy et al., 1969). This indicates that at least some of the carbonatites are cumulate igneous
rocks and, as shown in Afrikanda, genetically linked with silicate rocks (Chakhmouradian and
Zaitsev (in press). In the later evolution of the UACC, several stages of magmatic phoscorite-
carbonatite series were formed. The phoscorites are composed of magnetite, forsterite,
clinopyroxene, apatite and carbonate in different proportions, with magnetite usually dominant.
Calciocarbonatites of the first stages of these phoscorite-carbonatite series (Table 2) contain
forsterite, clinopyroxene, monticellite, phlogopite, amphibole, apatite and magnetite. Those
associated with the later stages are composed of calcite-dolomite and dolomite with forsterite and
tetraferriphlogopite (Table 2). The carbonatites of the latest stages are composed of Fe-carbonates
(Fe-rich dolomite, ankerite and siderite). They show pegmatitic textures and contain Ba, Sr, Nb and
REE mineralisation and quartz. They were formed at the end of the late-magmatic period and in the
post-magmatic one as pegmatitic and hydrothermal facies (Bulakh et al., 1998; Wall et al., 2001).
Geological, mineralogical, geochemical and isotopic evidence shows that there are genetic links
between phoscorite and carbonatite (Sokolov 1983; Balaganskaya, 1994; Zaitsev and Bell, 1995;
Subbotin and Subbotina, 2000; Krasnova et al., in press; Demeny et al., in press). The phoscorite-
carbonatite rock series represents the result of complex differentiation of a carbonate-silicate melt
that was extremely rich in iron and phosphorus. The early cumulates in each magma portion were
forsterite/diopside, magnetite and apatite (forming phoscorite), whereas the later ones were
dominated by carbonate minerals.

2.3 Alkaline syenite magmatism

This is represented by the Khibiny and Lovozero alkaline syenitic plutons (Kogarko, 1987),
the Kurga intrusion (Kogarko et al., 1995) and the Niva syenite intrusion (Arzamastsev et al.,
2000b). The volumetric importance of these massifs was shown by Arzamastsev et al. (2001) whose
detailed calculations suggest that the total volume of the two massifs formed 70% of the total
volume of magma in the KACP. The Khibiny pluton is the largest with an exposed area of 1327
km2. It is a multiphase intrusion, with a concentrically zoned structure formed by layered intrusive
bodies (Kukharenko et al., 1965). Its periphery consists of alkaline syenites, giving way to various
types of nepheline syenite (khibinites and rischorrites), followed by ijolite-melteigite-urtites and
foyaites towards the centre. A stock-like body of carbonatite cuts the foyaites (Dudkin et al., 1984;
Zaitsev, 1996). Contacts between intrusions are usually gradual. Only the ijolite-melteigite-urtite
intrusion has rather sharp contacts, showing complicated relations with the later intrusions (Ivanova
et al., 1970). Xenoliths of ultramafic and melilite-bearing rocks occur along its outer contacts
(Galakhov, 1975, 1988; Schpachenko and Stepanov, 1991), interpreted as relics of an earlier
ultramafic-alkaline intrusion (Dudkin et al., 1986; Balaganskaya and Savchenko, 1998). These
ijolites are similar in composition to the ijolite-urtite arc (Kostyleva-Labuntsova et al., 1978).
Numerous late-stage dykes of alkali picrite, monchiquite, nephelinite and phonolite cut the Khiniby
massif (Sindern et al., 2004).
The Lovozero multiphase nepheline-syenite pluton has an exposed area of 650 km2. It has
the form of a broad-based lopolith that can be traced geophysically down to 7 km (Schablinsky,
1963). Early nepheline and nosean syenites form about 5% of its area. The main pluton area (77%)
comprises a regular alternation of urtite, foyaite and lujavrite layers from a few centimeters to
hundreds of meters in thickness. Rocks of the eudialyte lujavrite suite represent a third phase and
form about 18% of the pluton area. Rare alkaline lamprophyre dykes cut all the earlier rocks
(Bussen and Sakharov, 1972; Kogarko et al., 1995; 2002).
The Kurga complex is a lopolith that can be traced to 4.5 km depth (Arzamastsev et al.,
1999). Its exposed area is about 30 km2 and is elliptical in shape. From the center to the margin, the
complex consists of cumulate olivinites, wehrlites and clinopyroxenites; the central part is cut by
stocks of alkaline and nepheline syenites (Kukharenko et al., 1971; Kogarko et al., 1995).

2.4 Alkali basaltic and carbonatitic extrusive rocks

Eruptive volcanic rocks are rare within the KACP, but xenoliths of alkali basaltic lavas up to
several hundred meters long occur within the Khibiny and Lovozero massifs (Gerasimovsky et al.,
1966; Galakhov, 1975; Bussen and Sakharov, 1972; Kogarko et al., 1995). These xenoliths are
thought to form part of the roof complex into which the plutonic rocks were intruded. A small
volcanic series interpreted as a carbonatite-melanephelinite palaeovolcano has been identified at
Kontozero (Pyatenko and Osokin, 1988). The total thickness preserved within a circular depression
in Kontozero is approximately 3km. Remnants of the lava pile have been found in Ivanovka on the
northern coast of the Kola Peninsula (Arzamastsev et al., 1998; 2002). They are nepheline basalts
and alkaline trachyte or phonolite lavas, up to 40m thick and extending for a few hundred meters.
Chemical analyses of the effusive silicate rocks are given by Arzamastsev et al (1998). Age
determinations (Table 3) suggest that they are 20-30 Ma older than the intrusive rocks, but
Kirichenko (1962) reported Upper Devonian – Lower Carboniferous plant fossils within the
sedimentary sequences accompanying the lavas in Lovozero and Kontozero. The original thickness
of the lava pile is one of the unknown factors in estimating the volume of magma that formed the
KACP, and the volume calculated by Arzamastsev et al. (2001) does not contain any value for
eroded extrusive rocks. A conservative estimate, based on a 2km thick pile of lavas covering an area
200km in diameter, yields an additional volume of 60000 km3 for the erupted material.
3. Geochronology
Rb-Sr dating of many of the alkaline complexes led Kramm et al. (1993) to the conclusion
that the KACP was formed during a relatively short time interval of 360 - 380 Ma. Further studies,
using techniques such as Ar-Ar, U-Pb and Sm-Nd, have generally confirmed this interval but have
also allowed it to be slightly broadened (Arzamastsev et al., 1999, 2002; Arsamastsev & Belyatsky,
2000). A recent review of the subject is given by Kramm and Sindern (in press) and our brief
discussion here generally agrees with these authors. Table 3 gives a review of the available age
determinations. Figure 2 shows a histogram of age determinations of KACP magmatic rocks,
derived from Table 3, in which the vast majority of the results still fall between 360 and 380 Ma.
The ages of the Kola kimberlites is poorly known, but one of the associated melilitites at Terskiy
Bereg has been dated at 365±16 Ma and a monticellite kimberlite dyke near Kandalaksha has
yielded an age of 382±14 Ma (Beard et al., 1998). However, Delenitsyn et al. (2001) determined
ages of 457±46 Ma (Sm-Nd isochron) and 465±12 Ma (Rb-Sr isochron) on separated minerals from
the Yermakovskaya-7 kimberlite pipe in Tersky Bereg. The Arkhangelsk kimberlites are also
constrained to be Devonian by fossil evidence (Sablukov, 1984; Makhotkin et al., 2000). Therefore
we consider that the kimberlite magmatism is an integral part of the KACP event.
A clinopyroxenite and an early carbonatite of the Seblyavr intrusion have yielded Rb-Sr
ages of 410±7 and 408±7 Ma (Balaganskaya et al., 1999b; Gogol & Delenitsyn, 1999), whilst
volcanic rocks of the Kontozero caldera and Lovozero were dated by Rb-Sr at 461±39 and 446±56
Ma, respectively (Arzamastsev et al., 1998). These older ages may represent earlier phases of
activity but it is possible that the samples may contain contaminants from the surrounding
Precambrian basement that have caused these older ages. Kramm and Sindern (in press) have
suggested that the Ar-Ar ages of 378-395 Ma for carbonatite and hornblendite xenoliths from the
Kandalaksha Gulf (Beard et al., 1996) may be due to excess argon. A much younger age (252 Ma)
has been obtained on a sample from Salmagora (Table 2) but this is considered to be spurious,
particularly as other samples from the same intrusion have yielded ages within the 360-380 Ma age
range (Table 3). Thus, further studies of the ages of KACP magmatism have largely supported the
assertion of Kramm et al. (1993) that the magmatic event was of relatively short duration.

4. Geochemistry
4.1 Major element geochemistry
Table 4 presents geochemical data for a selection of silicate dyke and pipe rocks of the
KACP, all of which are assumed to represent true magmatic liquids (i.e. have not undergone
cumulus processes). The loss on ignition (LOI) is included in the total in these analyses, so
recalculation to anhydrous compositions would raise the SiO2 content to some extent. Nevertheless,
on the Total Alkalis vs Silica diagram (Fig. 3) the dykes and pipes largely fall in the foidite field, as
they mostly have significantly lower SiO2 contents than common silica-undersaturated magmas
such as basanites and nephelinites. They also generally have low total-alkali contents (<5 wt%),
although some evolved melilitites and melanephelinites (< 7 wt% MgO) have higher total alkali
contents. These latter approach the compositions of nepheline syenites (Fig. 3). Within the dykes
and pipes, sodic types are more abundant than potassic ones (Fig. 4), with the potassic rocks tending
to be melilitites, damkjernites and kimberlites. Among the dyke rocks, CaO/Al2O3 ratios range from
1 to >5, with the higher values being from kimberlites and damkjernites, greatly exceeding the
ratios shown by typical ocean-island basaltic magmas and MORB (Fig. 5). This indicates that the
KACP magmas were derived from a mantle source with high CaO/Al2O3 ratios (e.g. one enriched in
carbonate).
The MgO-SiO2 diagram (Fig. 6) shows that many of the dyke rocks are highly magnesian
(>12 wt% MgO) and silica-undersaturated (20-40 wt% SiO2). When other rock-types such as
carbonatites, phoscorites, nepheline syenites and cumulates are plotted on the MgO-SiO2 diagram,
the cumulate rocks appear as mixtures of their constituent minerals (i.e. between diopside, olivine,
nepheline, calcite, dolomite, phlogopite and melilite). There are complete compositional trends
between carbonatites and olivinites, and between carbonatites and pyroxenites, suggesting that
many of the plutonic carbonates are cumulates. Phoscorites also plot between calcite and one or
more mafic silicate phases (forsterite, diopside and phlogopite) and may therefore be cumulates. In
constrast, the nepheline syenites form a fractionation trend to high SiO2 and low MgO, and overlap
with the more evolved foidolite dykes (Fig. 6).

4.2 REE and incompatible trace elements

Figure 7 shows chondrite-normalised REE abundances of primitive mafic alkaline rocks
(including kimberlites) from the KACP (Table 4). Their patterns are all remarkably similar and
show strong LREE enrichment with La/Yb between 70 and 200. Their LREE abundances are much
higher than those of OIB, indicating a more extreme metasomatism in the mantle source or a much
smaller degree of partial melting, whereas the HREE abundances are often lower than OIB,
suggesting the presence of residual garnet. Significantly, carbonatites from the KACP have
extremely similar REE patterns to those of the silicate magmas. This observation, however, does
not help in distinguishing the origin of the carbonatites.
The REE compositions of nepheline syenites from the Khibiny and Lovozero plutons
(Arzamastsev et al., 2001) show both higher and lower patterns than those of the supposed parental
magma of the UACCs (Fig. 8). This is probably related to the extensive fractional crystallisation
and accumulation that the evolved magmas have undergone, and the relative abundances of REE-
bearing minerals in the cumulate rocks. However, with the exception of a few rock-types which
show slight MREE depletion (probably due to titanite fractionation), the REE patterns in the
evolved rocks are parallel to those of the mafic and ultramafic parental magmas.
Mantle-normalised incompatible trace element diagrams for primitive silicate magma
compositions from the KACP (Fig. 9) show that they are enriched in strongly incompatible
elements relative to ocean island basalts (Sun & McDonough, 1989), although kimberlites and
melilitites show a trough at Zr. Both potassic and sodic magmas show very similar trace element
patterns. Such strong enrichment would be hard to explain even by small degrees of melting (0.1-
3%) of an undepleted garnet peridotite mantle. Most likely the source contains additional
metasomatic phases. Arzamastsev et al (2001) suggested the presence of phlogopite and amphibole,
together with possible apatite and perovskite, in the mantle source of the KACP magmas.
Figure 10 shows mantle-normalised trace element diagrams for carbonatite dykes from
Kola, i.e. those representing a magmatic liquid rather than a cumulate (Verhulst et al., 2000;
Dunworth and Bell, 2001; Mahotkin et al., 2003). They show distinct troughs at Rb and Nb,
although in all other trace elements they are identical to melilitites and/or kimberlites, including the
trough at Zr. However, some carbonatite dykes from Turiy (Ivanikov et al., 1998) are more enriched
in Sr and the REE, and do not show the same Nb trough (Fig. 10). Thus there appear to be at least
two “end-member” types of carbonatite magmas within the KACP, those with depletion in Nb and
those that show enrichment in this element. It is not clear what the petrogenetic significance of this
observation is, as these do not seem to correlate with any other feature (e.g. age, mode of
emplacement, etc).
When the trace element diagrams of nepheline syenites are compared with that of a typical
ultramafic lamprophyre or UACC parent magma (Fig. 11), many of the patterns are remarkably
similar, although some of the evolved rock-types actually have lower overall trace element contents.
This indicates that both crystal fractionation and crystal accumulation have occurred. Some highly
evolved nepheline syenites from Khibiny show extreme enrichment in the REE and Nb, but with
conspicuous troughs at Zr, Sr, Rb and Ba. We suggest that this pattern probably results from
fractionation of alkali feldspar and a Zr-bearing phase such as eudialyte.

4.3 Noble gas studies

More than 300 rocks and minerals from the KACP have been studied for noble gas
concentrations (He, Ne and Ar) and isotope systematics (Tolstikhin & Marty, 1998; Tolstikhin et
al., 1999, 2002). The highest 3He concentrations measured in any terrestrial sample, 4×10-9 cc/g,
with a rather low 4He/3He ratio of 3.3x104, were measured in a magnetite separate from a Seblyavr
clinopyroxenite (Kamensky et al., 2000). From the noble gas compositions, Tolstikhin et al. (2002)
considered that the KACP magmatism was related to a lower mantle plume with an initial 4He/3He
of 3 x 104. On the basis of the noble gas abundances in the principal terrestrial reservoirs
(Tolstikhin & Marty, 1998), the contribution of the lower mantle in KACP parental melts was
estimated to be 1.8 wt.%. This implies that heat from the plume caused melting of the upper mantle,
possibly including metasomatised subcontinental lithosphere (Tolstikhin et al., 1999, 2002).

4.4 Radiogenic isotope ratios

Kukharenko et al. (1965) and Galakhov (1975) first suggested a mantle origin for the
parental melts within the KACP from petrologic and geochemical data. This was later supported by
Sr and Nd radiogenic isotope studies (Kramm et al., 1993; Kramm, 1993; Kramm & Kogarko,
1994; Beard et al., 1996; 1998; Rukhlov, 1999; Verhulst et al., 2000; Dunworth and Bell, 2001).
Table 5 gives the data sources and values for the initial isotope ratios of radiogenic Sr and Nd, for a
wide variety of rocks from the KACP. Initial 87Sr/86Sr ratios at 380 Ma for intrusive silicate rocks
from the ultramafic alkaline complexes and the nepheline syenites vary from 0.7030 to 0.7049,
whilst the carbonatites show variations from 0.7030 to 0.7041. εNdt values (t = 380 Ma) determined
for intrusive silicate rocks in the KACP vary from +5.5 to +0.9, excluding an ijolite from Kovdor
with εNdt = -3.5; for Kola carbonatites εNdt is +7.3 to +0.4. Table 5 indicates that for some intrusions
such as Kovdor and Turiy, there is a large range of initial isotopic compositions, suggesting open
system behaviour, whereas in other intrusions (e.g. Vuoriyarvi) the isotopic compositions are much
more clustered, indicating a closed system for magma evolution.
Figure 12 shows the age-corrected Sr-Nd results for all analysed rocks from the KACP
using data from Table 5, Dunworth and Bell (2001) and Mahotkin et al. (2003). Most of them plot
in the Depleted Mantle field, with low 87Sr/86Sr and high εNdt values. They appear to diverge from a
depleted end-member that has been suggested to be the Devonian equivalent of FOZO (Bell and
Rukhlov, in press). The detailed variations in isotope compositions have previously been explained
by mixing of melts from two, three or even four mantle and crustal sources (Kramm, 1993; Kramm
et al., 1993; Zaitsev & Bell, 1995; Verhulst et al., 2000; Arzamastsev et al., 2000b; Dunworth &
Bell, 2001; Mahotkin et al., 2003; Sindern et al., 2004). Potassic samples are distinguished on Fig.
12 and most of them show either higher 87Sr/86Sri ratios or lower εNdt than the sodic rock-types.
Crustal contaminants could include the lower crust, represented by granulite xenoliths from
Kandalaksha (Kempton et al., 2001).

5. Discussion
5.1 General considerations
Igneous rocks of the KACP can be broadly classified into four groups: (a) primitive mantle-
derived silicate magmas, such as kimberlites, melilitites, nephelinites, ultramafic lamprophyres;
these are mostly found as dykes and pipes; (b) evolved alkaline and nepheline syenites found as
large plutons and as dykes cross-cutting the alkaline ultramafic complexes; (c) cumulate rocks,
including olivinites, clinopyroxenites, foidolites, melilitolites etc; these are found in the alkaline
ultramafic complexes and as xenoliths in nepheline syenites; (d) calcite carbonatites found within
the UACC (together with dolomitic carbonatites and phoscorites) and also forming dykes and
xenoliths. The relationship between all these rock-types can be investigated using the petrological
and geochemical data presented above. Borodin et al (1976) suggested that there were three parental
magmas within the KACP that had been derived from progressively deeper sources: alkali basaltic,
alkali ultramafic and kimberlitic.
One very important consideration about KACP rocks is that they show a very strong
tendency to low SiO2 contents, much lower than most common alkaline magmas, as shown in the
TAS diagram (Fig. 3). This indicates an extremely low degree of partial melting in the source
mantle. The abundance of magmatic carbonate suggests that the primary magmas were derived
from a carbonated mantle and the effect of the presence of carbonate in the mantle may enhance the
degree of silica undersaturation of the magmas formed from it. The KACP rocks also have much
more enriched REE patterns compared to OIB (Fig. 7), again indicating a lower degree of melting
than normal. The low HREE abundances of all primitive rock-types (Fig. 7) indicate the presence of
residual garnet, and the presence of diamonds in the Terskiy kimberlites (Kalinkin et al., 1993) are a
clear indication that at least some of the magmas were generated at depths >120km. It may be
significant that damkjernites and kimberlites show the lowest HREE contents (Fig. 7) and are
thought to come from the greatest depths.
Initial Sr and Nd isotope ratios of most of the rock-types (UACC, syenites, carbonatites) fall
in the Depleted Mantle field (Fig. 12) and the magmas must therefore have been derived from a
time-integrated Rb- and LREE-depleted mantle source. The metasomatic enrichment of the upper
mantle (deduced from the trace element enrichment) probably took place in Late Proterozoic or
Early Palaeozoic times, as suggested by Artemieva (2003). The spread of data requires slight
variations within this mantle source, with one end-member being rather depleted (εNdt = 6; 87Sr/86Sr
= 0.703) and another being closer to Bulk Earth (εNdt = 2; 87Sr/86Sr = 0.704). In contrast, many of
the pipes and dykes, particularly the potassic ones, have isotopic compositions that are quite
different from those of the majority of the KACP magmas, indicating either a source with a
significantly different isotopic composition or a different extent of alteration.
There is no obvious age progression for the various types of magmatic activity (Table 3;
Fig. 2). Thus, all of the magma types were formed in the same series of events over a relatively
short time span and over a relatively restricted area, and must therefore be related to some extent.
Recent experimental results on the melting of carbonated mantle at 6GPa (Dalton and Presnall,
1998) have shown that kimberlites and carbonatites can be formed from the same source during the
same melting event. These authors suggested that there is a strong petrogenetic link between
carbonatites, kimberlites and alkaline ultramafic rocks such as ultramafic lamprophyres. Rock
(1991) also pointed out that there is a compositional continuum between these magma types. The
noble gas systematics (Tolstikhin et al., 2002) suggest an origin related to mantle plume activity.
It is quite possible, therefore, that all of the magmatic rocks seen in the Devonian KACP
originated in a single event, triggered by the arrival of upwelling hot material (mantle plume?)
beneath the metasomatised Archaean/Proterozoic lithosphere of the Kola region. The influx of heat
caused melting of the carbonated garnet peridotite asthenospheric mantle, forming sodic magmas
and carbonatites, which evolved to form the UACC and the nepheline syenites. Pockets of
phlogopite-metasomatised lithospheric mantle were also melted and formed potassic ultramafic or
mafic magmas, which appear to have not evolved to form fractionated magmas. The depth and
degree of melting are probably the major factors controlling the precise composition of the melts
formed, together with variations in metasomatic phases in the mantle source.

5.2 Origin of silicate dyke rocks (magmatic liquids)

Borodin et al (1976) first suggested that there is a close petrogenetic relationship between
the dykes and the ultramafic alkaline complexes. Dyke rocks from the dyke swarms clearly
represent magma compositions, rather than the cumulate compositions that characterise the
ultramafic-alkaline magmatic complexes, and hence they can be used to investigate the primitive
magmas of the KACP. Table 4 shows some compositions of these rocks. All have high MgO (>12
wt%) and are clearly capable of being derived directly from the mantle. They frequently have
>500ppm Ni and >1000ppm Cr, supporting this suggestion. Ivanikov et al (1998) suggested that an
olivine melilite melanephelinite, derived from depths of at least 80km, was the primary magma for
the Turiy dyke series.
On the MgO-SiO2 diagram (Fig. 6), some of the primitive silicate dyke rocks show a
remarkable positive correlation between these oxides. This may relate to mixing between more and
less carbonated mantle source regions, or to olivine fractionation causing a decrease in MgO and
SiO2 in such silica-poor magmas. Damkjernites fall within this array, having low silica contents
(25-27 wt% SiO2) and only 10-12 wt% MgO (Table 4). Although they are potassic magmas, they
appear to be transitional to true carbonatites as they contain >20% CaO. CaO/Al2O3 ratios for
silicate dyke rocks range up to 5 (Fig. 5), indicating a very small degree of melting of a carbonated
mantle at depth. The damkjernites again have the highest CaO/Al2O3, transitional to silico-
carbonatites which have CaO/Al2O3 ratios >>7.
Mantle-normalised trace element diagrams for rocks from the ultramafic and mafic dykes
and pipes (Fig. 9) show that kimberlite, melilitite and damkjernite have the lowest HREE
abundances, indicating a lower degree of melting and greater depth of formation. Modelling by
Arzamastsev et al (2001) indicates that the spectrum of magmas from kimberlites to nephelinites
cannot be derived from small degrees of non-modal partial melting of an undepleted garnet
peridotite mantle source. However, the primitive magmas can be modelled as 0.3-0.5% melts of an
enriched phlogopite- and amphibole-bearing garnet lherzolite (Arzamastsev et al., 2001). Such
mantle rocks have been observed in the xenolith suite from the Arkhangelsk diamondiferous
kimberlites (Sablukov et al., 2000). Spinel peridotites from Kola also contain phlogopite, amphibole
and apatite (Arzamastsev and Dahlgren, 1994; Beard et al., in prep).
The Sr and Nd isotope ratios of the KACP silicate dyke rocks show a wide range of values
(Fig. 12). Many fall in the “Depleted Mantle” field, overlapping the values of KACP carbonatites,
UACC and syenites; however, numerous dykes have either much higher 87Sr/86Sr or much lower
εNdt values, similar to those found among the kimberlites and melilitites (Table 5). Several processes
could have given rise to these unusual isotopic ratios. One possibility is that they could have been
derived from the mantle lithosphere, in which either phlogopite or amphibole was present.
Phlogopite has a high Rb/Sr and over time would cause the region of the mantle in which it was
situated to have a high 87Sr/86Sr. However, mantle phlogopite has very low REE contents and so its
effect on the Nd isotope ratio would be limited. In contrast, mantle amphibole has modest Rb/Sr
ratios but is often highly LREE-enriched, which would cause low εNd values to evolve. The impact
of the mantle plume on the lithosphere may have caused these low-solidus domains to undergo
partial melting, thus producing a range of sodic and potassic magmas. An alternative possibility is
that all the samples with low εNdt values may be contaminated by continental crust, whereas those
with high 87Sr/86Sr may result from hydrothermal alteration that has affected Sr but not the REE.
Without detailed oxygen and Sr isotope measurements of individual phases within the samples, this
suggestion cannot be verified. However, it is supported by the leaching experiment of Beard et al.
(1996) in which the measured Sr isotope composition of a Kola ultramafic lamprophyre dyke was
reduced from 0.711 to 0.709 by leaching in hot acid.
Some melilitites and melanephelinites have low MgO contents (<10 wt%) and must
therefore be the products of fractionated magmas (Ivanikov et al., 1998). They are transitional in
their Si-Mg compositions to some of the rocks in the syenite intrusions (Fig. 6). They also have
much lower Ni and Cr contents (4-20ppm Cr; 35-65 ppm Ni), indicating that they have experienced
fractionation of mafic minerals. These samples, however, have Sr and Nd isotope compositions
identical to those of the more primitive magmas, demonstrating that fractionation occurred without
any contamination by the surrounding crust. Ivanikov et al. (1998) suggested that fractionation of
olivine, clinopyroxene, melilite, titanomagnetite, apatite and perovskite would be able to produce
the compositions of these more evolved magmas. They also pointed out that such a sequence of
crystallising phases would account for the compositions of many of the ultramafic rocks in the
UACC. Dunworth and Bell (2001) also suggested that there should be a large volume of olivine-
and pyroxene-rich ultramafic cumulates associated with fractional crystallisation within the Turiy
complex.

5.3 Origin of ultramafic alkaline carbonatite complexes

On the MgO-SiO2 diagram (Fig. 6), the cumulate rocks of the UACC plot as mixtures of
their constituent minerals. Together with the petrographical evidence for cumulus processes, this
indicates that few of the rocks found in the plutonic complexes are true magmatic liquids. Table 4
presents a calculated primitive magma composition for the ultramafic alkaline carbonatite
complexes (Arzamastsev et al., 2002). On the TAS diagram (Fig. 3), the proposed parental UACC
magmas is a typical silica-undersaturated KACP magma. It falls in the region of the MgO-SiO2
diagram (Fig. 6) occupied by melilitites and is within the compositional field of the melilitite and
ultramafic lamprophyres on the CaO/Al2O3 vs SiO2 diagram (Fig. 5). Furthermore, the parental
UACC magma has a REE pattern identical to that of melilitites and carbonatites from the KACP
(Fig. 7). It has high Ni (242 ppm) and Cr (393 ppm) contents, again indicating mantle derivation. In
terms of their radiogenic isotope ratios, UACC silicate rocks largely overlap the field of
carbonatites, syenites and many of the dyke rocks, and so must have been derived from the same
mantle source. Crustal contamination may have occurred in a single sample of ijolite from Kovdor
that has an unusually low εNdt ratio (Verhulst et al., 2000), but this needs to be confirmed by oxygen
isotope analysis.

5.4 Origin of the alkaline syenites

Kogarko (1977a, 1977b, 1979) discussed the origin of nepheline syenites in the KACP and
used phase diagram constraints to show that the nepheline syenite minimum composition can
appear as a differentiation product from a variety of nepheline-normative parental melts, such as
olivine nephelinites, melilitic basalts and alkali basalts. Typical compositions of the nepheline
syenites in the Khibiny and Lovozero plutons (Arzamastsev et al., 2001) have low MgO contents
(<2 wt% MgO) and extreme enrichment in alkalis (Na2O > 8 wt%; K2O > 3.5 wt%). On the TAS
diagram (Fig. 3), they plot in the evolved foidite field that has even less SiO2 than common
nepheline syenites. Such extreme compositions could not be derived directly from the mantle, so it
is clear that the magmas in these intrusions have undergone extensive fractionation. Ultramafic
cumulate rocks are present in the Khibiny and Kurga intrusions, indicating a close relationship.
Much more ultramafic rock must be present at depth within the crust, formed of minerals that have
been crystallised during this fractionation process. Arzamastsev et al. (2001), using density
modelling and borehole core, suggest that about 50% of Khibiny is formed of alkaline ultramafic
rocks. The REE pattern of the most evolved nepheline syenites (Fig. 8) is more than an order of
magnitude more enriched than that of a typical ultramafic parental magma, suggesting extreme
fractionation.
Table 1 shows that syenites are present in more than 50% of the UACC, indicating that such
magmas were formed during the evolution of the UACC. During its late magmatic stage, the
Kovdor UACC pluton was intruded by dykes of alkaline syenite and nepheline syenite. A xenolith
of layered eudialyte syenite found in a dyke in the Turiy area (Beard et al., in prep) indicates the
existence of an agpaitic syenitic body at depth, thus forming another link between the UACC and
the syenitic plutonic magmatism. On the MgO-SiO2 diagram (Fig. 6) there is an overlap between
the least evolved nepheline syenites and the most evolved foidolites. The similarity of their Sr and
Nd radiogenic isotope ratios to those of more mafic rocks in the UACC (Fig. 12) also supports the
suggestion that fractional crystallisation was the main process that formed these rocks. Minerals
that would fractionate from silica-undersaturated mafic alkaline magmas such as nephelinites
include olivine, clinopyroxene, nepheline, melilitite and eventually alkali feldspar (Dubrovsky,
1989). These minerals are the most common constituents of the cumulate rocks in the UACC and in
the ijolite-urtite arc within the Khibiny massif.

5.5 Origin of potassic magma types, including kimberlites

Figure 4 shows that there is a continuum of sodic and potassic magma types within the
dykes and pipes. The most common potassic magma type is olivine phlogopite melilitite, although
damkjernites and kimberlites are also present. Potassic magmas clearly have a different mantle
source mineralogy than sodic magmas and derivation from a phlogopite-bearing lithospheric mantle
is one possible explanation. Nevertheless, their REE and mantle-normalised trace element patterns
are very similar to those of the sodic magmatic rocks, indicating a broadly similar source (Figs. 7
and 9). Although some potassic magmas have Sr and Nd isotope compositions overlapping with the
main field of KACP magmatism (Fig. 12), many show much higher 87Sr/86Sr or lower εNdt values.
These unusual values may be derived from the enriched mantle lithosphere. Such lithosphere may
be represented by phlogopite clinopyroxenites that occur with garnet peridotite xenoliths in the
Arkhangelsk kimberlites. Alternative explanations could be that these rocks are either
hydrothermally altered or contaminated by the Archaean/Proterozoic crust. Beard et al (1998; 2000)
discussed the origin of kimberlites in Terskiy Bereg and Arkhangelsk, and concluded that they were
formed from a metasomatised mantle source and were associated with impact of a mantle plume on
the base of the thick cratonic lithosphere.

5.6 Origin of carbonatites and phoscorites

The evolution and source characteristics of carbonatite magmas in the KACP have been
recently reviewed by Bell and Rukhlov (in press). There is a long-standing debate about whether
carbonatite magmas are products of liquid immiscibility or can be derived directly from melting of
the mantle (or both). However, many recent experimental studies (e.g. Wallace and Green, 1988;
Thibault et al., 1992; Dalton and Wood, 1993; Lee and Wyllie, 2000) have demonstrated that
carbonatite melts can be formed from carbonated lherzolite at moderate to high pressures. The
presence of carbonatites as dykes and as volcanic products (e.g. in Kontozero) indicates that
carbonate magmas were present as separate entities in the KACP, but it is not clear whether they
were formed separately at mantle depths or exsolved at lower pressures. Evidence for the passage of
carbonatite melts through the lithospheric mantle is suggested by the presence of wehrlite xenoliths
associated with spinel peridotite xenoliths in a dyke near Kandalaksha (Beard et al., in prep).
Wehrlites can be formed by carbonatite metasomatism within the mantle.
The similarities of the REE and trace element patterns for ultramafic alkaline silicate rocks
and carbonatite dykes indicate that there is a strong relationship between the two magma types. This
is seen in the field relations within the UACC, in which carbonatite rocks grade into silicate rocks.
Bell and Rukhlov (in press) point out that fractional crystallisation may be the dominant process in
the origin of carbonatites. Primary igneous calcite is common in many silicate rocks within the
KACP, e.g. as interstitial phase in the Kovdor ultramafic cumulates (Verhulst et al., 2000), and in
dykes associated with the Kandaguba complex (Pilipiuk et al., 2001) and Pinozero dyke swarm
(Rukhlov, 1999; Bell and Rukhlov, in press). Many of the carbonatite rocks in the UACC are in fact
cumulates, so their compositions represent a mixture between cumulus carbonate minerals and
cumulus silicate minerals (Fig. 6). Plutonic silico-carbonatite xenoliths from a ferrocarbonatite sill
in the Kandalaksha Gulf were also interpreted as cumulates (Beard et al., 1996). Veksler et al.
(1998) analysed melt inclusions in silicate minerals from carbonatite and silicate rocks of Kovdor,
and suggested that fractionation of amphibole and phlogopite was an important process during melt
evolution. Many olivine-free ultramafic cumulate xenoliths have been found together with the
plutonic carbonatite xenoliths in Kandalaksha Gulf (Beard et al., in prep). The mineralogy of these
xenoliths includes pargasite, phlogopite, diopside and apatite; they show clear evidence of
magmatic layering.
In contrast to the suggestion that the carbonatite magmas were formed directly from the
mantle or from crystal fractionation, Bulakh and Ivanikov (1984) and Ivanikov et al (1998) argued
for an origin due to liquid immiscibility from nepheline melilitite magmas. They suggested that the
presence of carbonate ocelli in the mafic silicate dykes was consistent with this process. On a large
scale, the carbonatites and silicate rocks in the UACC also have similar Sr and Nd isotope ratios
(Fig. 12), although detailed studies of Rb-Sr and Sm-Nd systems suggest slightly different sources
for carbonatites and silicate rocks of the Khibiny complex (Kramm & Kogarko, 1994; Zaitsev et al.,
1997; Sindern et al., 2004). Detailed isotopic studies have shown that KACP carbonatites do not
have a uniform isotopic composition and therefore are probably derived from a heterogeneous
mantle source (e.g. Verhulst et al., 2000; Dunworth and Bell, 2001). Sr-Nd isotopic data cannot
easily distinguish between the two processes of partial melting or liquid immiscibility and, as Bell
and Rukhlov (in press) have concluded, there is no overall agreement about whether carbonatites
arrive in the crust as carbonate magmas or separate by liquid immiscibility at shallower levels.
Possibly both mechanisms have operated, even within the same complex.
 Phoscorites are extremely unusual rock-types, associated with carbonatites within the
UACC (Krasnova et al., in press). Their mineralogy is highly variable, ranging from almost
monomineralic to magnetite-apatite-forsterite/diopside rocks. Such mineralogical variations imply
that crystal segregation processes were involved in their formation (i.e. some phoscorites are
probably cumulates), but Krasnova et al. (in press) also suggest that liquid immiscibility between
phoscorite and carbonatite melts may have played a role. A phoscorite melt would be extremely
rich in Fe and P, and would have a very different melt structure from that of a carbonatite melt.
Stable isotope data for Kola carbonatite and phoscorite complexes have been presented by Demeny
et al. (in press), who suggested that liquid immiscibility may have operated to form some
phoscorites (e.g. in Sokli) but the involvement of separate sources is required in other examples
(e.g. Turiy).

5.7 Asthenospheric or lithospheric source?

In common with much of the Baltic Shield, the present-day lithospheric thickness beneath
the Kola Peninsula is at least 200km (Artemieva, 2003). The presence of rare diamonds in
kimberlites of the Terskiy coast (Kalinkin et al., 1993) and Arkhangelsk (Mahotkin et al., 2000)
suggests that lithospheric thicknesses were similar in Devonian times. This gives rise to questions
concerning the location of the melting that gave rise to the suite of magmas. The general
homogeneity in isotopic and REE composition within the non-potassic silicate magmas and the
carbonatites argues for a homogeneous source, which is conventionally considered to be the
asthenosphere, whereas the isotopic heterogeneity of the potassic magmas (including kimberlites)
suggests a lithospheric source. However, kimberlites are usually considered to be formed at greater
depths than melilitites, nephelinites or carbonatites. Thus, possibly all KACP magmas were derived
from a lithospheric mantle that had been metasomatised prior to melting, which had not had time to
develop extensive isotopic heterogeneity except in the regions containing high Rb/Sr phlogopite.
The kimberlites were derived from the deepest levels (at least 120 km), which were most
metasomatised and heterogeneous, with the nephelinites and melilitites being formed from
shallower and more homogeneous regions of the lithosphere. Re-activation of deep lithospheric
fractures may have contributed to the uprise of magma.
The trigger for the magmatism is also debateable. Numerous authors have suggested the
presence of a mantle plume (e.g. Beard et al., 1998; Dunworth and Bell, 2001; Sindern et al., 2004;
Bell and Rukhlov, in press). This would help to explain (a) the short time duration of the
magmatism and (b) the large volume of magma produced (between 15000 and 75000 km3,
depending on the original thickness and extent of erupted lavas). The highly silica-undersaturated
nature of the magmatism is, however, rather unusual for a plume event and tends to suggest very
small degrees of melting of an enriched source, rather than extensive melting related to the impact
of a large plume head. Even the He isotopic systematics suggests only a limited role for material
actually derived from the lower mantle (Tolstikhin et al., 1999, 2002). Moreover, Meiboom et al
(2003) have suggested that high 3He/4He ratios (= low 4He/3He ratios) may not be a reliable
indicator of the presence of a deep «plume» component.

6. Conclusions
From the data reviewed above, we conclude that most of the magmas within the Kola
alkaline province were derived from an upper mantle source (possibly a thick lithosphere), although
a lower mantle plume may have initiated the magmatism. A spectrum of magmas from carbonatite
to damkjernite to melilitite to ultramafic lamprophyre was derived from melting of metasomatised
carbonated garnet peridotite. Most primitive silicate magmas have very similar trace element and
isotopic compositions and therefore are derived from a similar mantle source. The most likely
'primitive magma' for the ultramafic-alkaline-carbonatite complexes is a sodic melilitite, olivine
melanephelinite or ultramafic lamprophyre. Depth of melting, degree of melting and mineralogy of
the mantle source are the key factors in determining the composition of the ultramafic and mafic
magmas. Potassic magmas such as kimberlites may have been derived from enriched mantle
lithosphere containing residual phlogopite. Cumulate processes were very important in forming the
spectrum of alkaline ultramafic rock-types in the ultramafic-alkaline-carbonatite complexes; crystal
fractionation processes were important in forming the nepheline and alkaline syenites. Carbonatites
may have been formed by direct mantle melting or by liquid immisicibility, but most of them are
probably related to crystal fractionation of carbonated alkaline magmas.

7. Acknowledgements
We are very grateful to Lia Kogarko and Frances Wall for their careful reviews of this
paper. We wish to record our thanks to the organisers of the annual SVEKALAPKO meetings, held
as part of the ESF-funded EUROPROBE project, that enabled Russian and Western colleagues to
exchange samples, data and ideas in the congenial surroundings of Lammi (Finland) and Repino
(Russia).

List of Tables
Table 1. Rock-types occurring in ultramafic-alkaline carbonatite complexes, KACP
Table 2. General scheme of evolution of the Kovdor ultramafic-alkaline carbonatite complex (after
Liferovich, 1998).
Table 3. Compilation of age determinations KACP
Table 4. Primitive silicate magma compositions from KACP
Table 5. 87Sr/86Sr(initial) and εNd(I) data for KACP, calculated to 380 Ma.
Appendix: Explanation of unusual rock-types found in the KACP

List of Figures
Fig. 1. Sketch map of Kola peninsula and NE Finland, showing location of main complexes and
dyke swarms / clusters of pipes of the KACP.
Figure 2. Age histogram of magmatic activity in KACP, derived from data and references in Table
3.
Figure 3. Wt. % Na2O + K2O vs SiO2 diagram for KACP dyke rocks, kimberlites, nepheline
syenites, and UACC parental magma. Data sources: Beard et al. (1996); Beard et al. (1998);
Arzamastsev et al. (1998); Beard et al. (2000); Arzamastsev et al. (2001); Sindern et al.
(2004); Mahotkin and Downes (unpub).
Figure 4. Wt. % K2O vs Wt. % Na2O diagram for rocks from KACP dykes and pipes, and parental
magma for UACC. Data sources as for Fig. 3.
Figure 5. CaO/Al2O3 vs Wt. % SiO2 for rocks from KACP dykes and pipes, and parental magma for
UACC. Data sources as for Fig. 3.
Figure 6. Variation diagram of Wt. % MgO vs Wt. % SiO2 for all varieties of KACP magmatic
rocks. Mineral abbreviations: Cc = calcite; Dol = dolomite; Mel = melilite; Fo = forsterite;
Di = diopside; Ne = nepheline; Phl = phlogopite.
Figure 7. REE plots of primitive magma compositions and carbonatite from KACP. Chondrite
normalisation coefficients from Nakamura (1974). Data sources: Table 4.
Figure 8. REE plots of evolved rock-types from KACP. Chondrite normalisation coefficients from
Nakamura (1974). Data source: Arzamastsev et al. (2001).
Figure 9. Trace element data for primitive magma compositions from KACP, normalised to
primitive mantle composition of Sun and McDonough (1989). Data sources: Table 4.
Figure 10. Trace element data for carbonatites from KACP, normalised to primitive mantle
composition of Sun and McDonough (1989). Data sources: Beard et al. (1996); Ivanikov et
al. (1998); Dunworth and Bell (2001).
Figure 11. Trace element data for evolved rock-types from KACP, normalised to primitive mantle
composition of Sun and McDonough (1989). Data source: Arzamastsev et al. (2001).
Figure 12. Sr-Nd isotope diagram for KACP rocks. Data sources: Table 5; Dunworth and Bell
(2001); Mahotkin and Downes (unpub).
References

Amelin Yu. V. and Zaitsev A., 1997. Precise U-Th-Pb chronology of carbonatites and phoscorites:
problems related to extreme elemental fractionation and possible solutions using multi-
mineral approach. GAC/MAC Annual Meeting Ottawa, Abstr. p2.
Artemieva I. M., 2003. Lithospheric structure, composition and thermal regime of the East
European Craton: implications for the subsiudence of the Russian Platform. Earth Planet.
Sci. Lett. 213, 431-446.
Arzamastsev, A.A., 1994. Unique Palaeozoic intrusions of the Kola Peninsula. Kola Science Centre
Press House, Apatity. 78pp (in Russian and English).
Arzamastsev A.A. and Dahlgren S., 1994. Plutonic mineral assemblages in Palaeozoic dikes and
explosion pipes of the Alkaline Province of the Baltic Shield. Geochem. Int., 31 (3) 57-68.
Arzamastsev, A.A., Arzamastseva, L.V. and Belyatsky B.V., 1998. Initial alkaline volcanism of the
Palaeozoic tectonic and magmatic activation in the North-Easten Fennoscandia:
geochemical features and petrological consequences. Petrology, 6, 316-336 (in Russian).
Arzamastsev, A.A. Bayanova T.B., Arzamastseva L.V., Balashov Yu.D. and Gogol O.V., 1999.
Initial magmatism of Palaeozoic tectono-magmatic activation in the northeastern Baltic
Shield: age and geochemical features of the Kurga massif, Kola Peninsula. Geochimia, 11,
1139-1150.
Arzamastsev, A.A., Glaznev, V.N., Arzamastseva, L.V. and Raevsky F.B., 2000a. The morphology
and internal structure of the Kola alkaline intrusions, NE Fennoscandian shield: 3D density
modelling and geological implications. Journal of Asian Earth Sciences, 18, 213-228.
Arzamastsev, A.A., Belyatsky, B.V. and Arzamastseva, L.V., 2000b. Agpaitic magmatism in the
northeastern Baltic Shield: a study of the Niva intrusion, Kola Peninsula, Russia. Lithos, 51,
27-46.
Arzamastsev, A.A. and Belyatsky B.V., 2000. Palaeozoic activation in the northeastern
Fennoscandian Shield: Rb-Sr and Sm-Nd isochron dating of initial volcanics and final dyke
pulses of magmatism. Abstracts of the SVEKALAPKO 5th workshop. University of Oulu,
Department of Geophysics, Report No. 23, p. 6.
Arzamastsev A.A., Bea F., Glaznev, V.N., Arzamastseva, L.V. and Montero P., 2001. Kola alkaline
province in the Palaeozoic: evaluation of primary mantle magma composition and magma
generation conditions. Russian Journal of Earth Sciences, 3, 1-32.
Arzamastsev A.A., Bea F., Belyatsky B.V., Glaznev V.N., Arzamastseva, L.V., Travin A. and
Montero P., 2002. Palaeozoic processes of plume-lithosphere interaction in the Northeast
Baltic Shield: prolongation, volumes, conditions of magma generation. In: F.P. Mitrofanov
(Ed.) Geology and raw materials of the Kola Peninsula. Apatity, KSC RAS. Volume 2. 104-
145 (in Russian).
Balaganskaya E.G., 1994. Breccias of the Kovdor phoscorite-carbonatite deposit of magnetite and
their geological meaning. Zapiski Vserossiuskogo Mineralogicheskogo Obschestva, 2, 21-
36. (In Russian).
Balaganskaya E.G. and Savchenko E.A, 1998. Perovskite in the rocks of the ijolite-urtite arc of the
Khibiny massif. Zapiski Vserossiuskogo Mineralogicheskogo Obschestva, 1, 85-96. (In
Russian).
Balaganskaya E., Downes, H., Subbotin, V., Liferovich, R. and Beard, A. 1999a. Mantle sources of
Kola carbonatites: new Sr, Nd and geochemical evidence from the Vuorijarvi Massif.
Abstracts of the 4th Svekalapko Workshop. University of Oulu, Department of Geophysics,
Rep. 1999. 22: 7.
Balaganskaya, E.G., Gogol, O.V., Liferovich, R.P., Delenitsin, A.A., Sherstobitova, G.M. and
Sherstennikova, O.G., 1999b. Origin of clinopyroxenite, phoscorite and carbonatite of the
Palaeozoic Seblyavr massif, north-western Russia: evidence from Rb-Sr data. J. Conf.
Abstr. 4(1): p350.
Balaganskaya, E., Verhulst A., Downes, H., Liferovich, R., Demaiffe D. and Laajoki K., 2000.
Geochemistry, petrography and mineralogy of clinopyroxenite, phoscorites and carbonatites
of the Seblyavr massif, Kola Alkaline Carbonatite Province, Russia Abstracts of the 5th
Svekalapko Workshop. University of Oulu, Department of Geophysics, Rep. 23, p8.
Balaganskaya, E.G., Downes H., Liferovich, R.P., Verhulst A., Demaiffe D. and Laajoki K. (in
prep.). Clinopyroxenite, phoscorites and carbonatites of the Seblyavr massif, Kola Alkaline
Province, Russia: mineralogy, geochemistry and origin.
Balagansky, V.V., Glaznev, V.N. and Osipenko, L.G., 1998. Early Proterozoic Evolution of the
Northeastern Baltic Shield: A Terrane Analysis. Geotectonics, 32, 81-92.
Bayanova T.B., Kirnarsky Yu. M. and Levkovich N.V., 1997. A U-Pb study of baddeleyite from
rocks of the Kovdor massif. Doklady Akadem. Nauk. 356, 509-511 (in Russian).
Beard A.D., Downes H., Vetrin V., Kempton P.D. and Maluski H. 1996. Petrogenesis of Devonian
lamprophyre and carbonatite minor intrusions, Kandalaksha Gulf (Kola Peninsula). Lithos,
39, 93-119
Beard, A.D., Downes, H., Hegner, E., Sablukov, S.M., Vetrin, V. and Balogh, K. 1998. Mineralogy
and geochemistry of Devonian ultramafic minor intrusions of the southern Kola Peninsula,
Russia: implications for the petrogenesis of kimberlites and melilitites. Contrib. Mineral.
Petrol., 130, 288-303.
Beard, A.D., Downes, H., Hegner, E., Sablukov, S.M., 2000. Geochemistry and mineralogy of
kimberlites from the Arkhangelsk Region, NW Russia: evidence for transitional kimberlite
magma types. Lithos 51, 47-73.
Bell, K. and Rukhlov A. S. in press. Carbonatites from the Kola Alkaline Province: Origin,
evolution and source characteristics. In: Phoscorites and Carbonatites from mantle to mine.
(Eds: F Wall and A N Zaitsev). Mineralogical Society Series 10.
Borodin L S., Lapin A. V. and Pyatenko I. K., 1976. Petrology and geochemistry of alkaline-
ultramafic dykes and kimberlites. Nauka, Moscow. 244p (in Russian).
Brassinnes S., Demaiffe D., Balaganskaya E. and Downes H., 2003. New mineralogical and
geochemical data on the Vuorijarvi ultramafic, alkaline and carbonatite complex (Kola
region, NW Russia). Periodico di Mineralogia, Vol LXXII, Special Issue No 1, 79-86.
Brassinnes S., Balaganskaya E.and Demaiffe D., in press. Magmatic evolution of the differentiated
ultramafic, alkaline and carbonatite intrusion of Vuoriyarvi (Kola Peninsula, Russia). A LA-
ICP-MS study of apatite. Lithos (this volume).
Bulakh A.G. and Ivanikov V. V., 1984. The Problems of Mineralogy and Petrology of Carbonatites.
L G U, Leningrad. 242p (in Russian).
Bulakh, A.G., Le Bas, M.J., Wall, F. and Zaitsev, A.N., 1998. Ancylite-bearing carbonatites of the
Seblyavr massif, Kola Peninsula, Russia. N. Jb. Mineral Mh., 4, 171-192.
Bulakh, A.G., Ivanikov V. V. and Orlova M. P., in press. Overview of carbonatite-phoscorite
complexes of the Kola Alkaline Province in the context of a Scandinavian North Atlantic
Alkaline Province. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F Wall and
A N Zaitsev). Mineralogical Society Series 10.
Bussen, I. V. and Sakharov, A. S., 1972. Petrology of the Lovozero alkaline massif. Nauka
Leningrad. 296p (in Russian).
Chakhmouradian A. R. and Zaitsev A. N. in press. Afrikanda: An association of ultramafic, alkaline
and alkali-silica-rich carbonatitic rocks from mantle-derived melts. In: Phoscorites and
Carbonatites from mantle to mine. (Eds: F Wall and A N Zaitsev). Mineralogical Society
Series 10.
Dalton J. A. and Wood B. J., 1993. The composition of primary carbonate melts and their evolution
through wall-rock reaction in the mantle. Earth Planet. Sci. Lett. 119, 511-525.
Dalton J. A. and Presnall D. C., 1998. The continuum of primary carbonatitic-kimberlitic melt
compositions in equilibrium with lherzolite: data from the system CaO-MgO-Al2O3-SiO2-
CO2 at 6 GPa. J. Petrol. 39, 1953-1964.
Delenitsyn A.A, Gavrilenko B. V., Serov P.N. and Bayanova T.B, 2001. Sm-Nd and Rb-Sr age
determination of Yermakovskaya-7 kimberlite pipe, Tyersky Bereg, Kola region. In:
SVEKALAPKO 6th workshop abstracts. University of Oulu, Department of Geophysics,
Report No. 24, p17.
Demeny A., Sitnikova M. A. and Karchevsky P. I., in press. Stable C and O isotope compositions of
carbonatite complexes of the Kola Alkaline Province: phoscorite-carbonatite relationships
and source compositions. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F
Wall and A N Zaitsev). Mineralogical Society Series 10.
Downes H., Peltonen P., Mänttäri I. and Sharkov E. V., 2000. Proterozoic zircon ages from lower
crustal granulite xenoliths, Kola Peninsula, Russia: evidence for crustal growth and
reworking. J. Geol. Soc. Lond. 159, 485-488.
Dubrovsky M. I., 1989. Phase diagram of part of the system larnite-forsterite-nepheline-quartz
(extended basalt tetrahedron) at 1atm. Apatity: Kola Science Centre Publication. 34p (in
Russian)
Dudkin O.B., Balaganskaya E.G. and Arzamastseva L.V., 1986. On the problem of the Khibiny
deposits formation. In: Deposits of the nonmetallic raw materials of the Kola Peninsula.
Published by Kola Branch of the USSR Acad of Sci. Apatity. P. 4-10.
Dudkin O.B., Minakov F.V., Kravchenko M.P., Kravchenko A.V., Kulakov A.N., Polezhaeva L.I.,
Pripachkin V.A., Pushkarev Yu.D. and Rungenen G.I., 1984. Carbonatites of Khibiny.
Printed by Kola Branch of the USSSR Academy of Sci. Apatity. 98 p.
Dunworth, E.A., 1997. The Turiy Massif, Kola Peninsula, Russia: Open-system disequil;ibrium.
PhD thesis, Carleton University, Ottawa, Ontario, Canada, 488pp.
Dunworth, E.A. and Bell, K., 2001. The Turiy Massif, Kola Peninsula, Russia: Isotopic and
Geochemical Evidence for Multi-source Evolution. J. Petrol., 42, 377-405.
Galakhov, A.V., 1975. Petrology of the Khibiny alkaline massif. Leningrad, Nauka Publishing
House, 256 p. (in Russian).
Galakhov, A.V., 1988. Khibiny massif – complicated poly-source intrusive of the central type.
Doklady AN SSSR. V. 302. No 3. 673-675. (In Russian).
Gerasimovsky V. I., Volkov V. P., Kogarko L. N., Polyakov A. I., Saprykina T. V. and Balashov
Yu. A., 1966. The geochemistry of the Lovozero alkaline massif. Part 1 and Part 2.
Australian National University press, Canberra.
Gogol O. V., Bayanova T. B., Balaganskaya E. G. and Delenistin A. A., 1998. New evidence of the
duration of alkaline magmatism of the Kola region (Russia) based on Rb-Sr and U-Pb
isotope data. Chin. Sci. Bull. 43, p45.
Gogol O. V. and Delenistin A A., 1999. New Rb-Sr data for Kola alkaline province. Proceedings of
the 10th Kratz conference. Apatity, Kola Sci. Centre Publ. 43-47 (in Russian)
Gogol O. V., Bayanova T. B. and Delenistin A A., 2000. Stages of the carbonatite formation on the
sample of the Seblyavr massif (according to isotope data). Proceedings of the XIX seminar
“Geochemistry of the magmatic rocks”, Moscow. GEOKHI Publ. 45-46 (in Russian).
Ivanikov V. V., 1977. Geology and petrology of the compex of alkaline ultrabasic dykes of the
Kandalaksha graben. PhD thesis. Leningrad University. 264p (in Russian).
Ivanikov V. V., Rukhlov A. S. and Bell K., 1998. Magmatic evolution of the melilitite-carbonatite-
nephelinite dyke series of the Turiy peninsula (Kandalaksha Bay, White Sea, Russia). J
Petrol. 39, 2043-2059.
Ivanova T.N., Dudkin O.B., Kosyreva L.V.and Polyakov K.I., 1970. Ijolite-urtites of the Khibiny
Tundras. Leningrad, Nauka Publish. House. 179 p. (in Russian).
Kalinkin M. M., Arzamastsev A. A. and Polyakov I. V., 1993. Kimberlites and related rocks of the
Kola region. Petrology 1, 173-180.
Kamensky I.L., Ikorsky S.V., Charkov I.V., Gannibal M.A., Novikov D.D. and Tolstikhin I.L.,
2000. Low mantle helium in Palaeozoic intrusions of Kola Peninsula. I Russian conference
on isotopic geochronology. Moscow, 15-17 Nov.2000. Abstracts vol.: “Isotopic dating of
geological processes: a new methods and results. P.165-167. (in Russian).
Karchevsky P. I. and Moutte J. in press. The phoscorite-carbonatite complex of Vuoriyarvi, Kola
Peninsula. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F Wall and A N
Zaitsev). Mineralogical Society Series 10.
Karchevsky P. I. in press. Turiy Mys phoscorite-carbonatite complex and its diopside-bearing
phoscorite. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F Wall and A N
Zaitsev). Mineralogical Society Series 10.
Kelley S. P. and Wartho J.-A., 2000. Rapid kimberlite ascent and the significance of Ar-Ar ages in
xenolith phlogopites. Science 289, 609-611.
Kempton, P.D., Downes, H., Neymark, L.A., Wartho, J.A., Zartman R.E. and Sharkov, E.V., 2001.
Garnet granulite xenoliths from the northern Baltic Shield - the underplated lower crust of a
Palaeoproterozoic Large Igneous Province? J. Petrol., 42, 731-763.
Kirichenko A A., 1962. Palaeozoic volcanic and sedimentary rocks of the Kola Peninsula. In:
Materials on the Geology and Minerals of northwestern RSFSR, 3, 27-44. Gostopttekhisdat,
Leningrad-Moscow (in Russian).
Kogarko, L.N., 1977a. Genetic problems of agpaitic magmas. Nauka Moscow, 294p (in Russian)
Kogarko, L.N., 1977b. General regularities of differentiation of magmas oversaturated with alkalis.
Geochemistry International 14, 9-25.
Kogarko, L.N., 1979. Microcomponents as indicators of the differentiation of alkaline magma
series, In: “Origin and Distribution of the Elements” (ed. L H Ahrens). Pergamon Press,
Oxford and New York.
Kogarko, L.N., 1987. Alkaline igneous rocks of the Kola Peninsula. In: Alkaline Igneous Rocks.
Geol. Soc. Sp. Pub. 30, Fitton J G and Upton BG J (Eds). 531-544.
Kogarko, L.N., Kononova, V.A., Orlova, M.P. and Woolley, A.R., 1995. Alkaline rocks and
carbonatites of the world. Part 2. Former USSR. Chapman and Hall, London, 225 p.
Kogarko, L.N., Williams C.T. and Woolley, A.R., 2002. Chemical evolution and petrogenetic
implications of loparite in layered, agpaitic Lovozero complex, Kola Peninsula, Russia.
Mineral. and Petrol. 74, 1-24.
Kostyleva-Labuntsova E. E., Borutsky B. E., Sokolova M. N., Shlukova Z. D., Dorfman M. D.,
Dudkin O. B., Kozyreva L. V. and Ikorsky S. V., 1978. Mineralogy of the Khibina Massif.
Nauka, Moscow. 228 p (in Russian).
Kramm, U. and Kogarko, L.N., 1994. Nd and Sr isotope signatures of the Khibiny and Lovozero
agpaitic centres, Kola Alkaline Province, Russia. Lithos, 32, 225-242.
Kramm, U., 1993. Mantle components of carbonatites from the Kola Alkaline Province, Russia and
Finland: A Nd-Sr study. Eur. J. Mineral., 5, 985-989.
Kramm, U., Kogarko, L.N., Kononova, V.A. and Vartiainen, H., 1993. The Kola Alkaline Province
of the CIS and Finland: Precise Rb-Sr ages define 380-360 Ma age range for all magmatism.
Lithos, 30, 33-44.
Kramm, U. and Sindern S. in press. Timing of Kola ultrabasic alkaline, alkaline and phoscorite-
carbonatite magmatism. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F Wall
and A N Zaitsev). Mineralogical Society Series 10.
Krasnova N.I., 2001. The Kovdor phlogopite deposit, Kola Peninsula, Russia. Can. Miner. 39, 33-
44.
Krasnova N.I. and Kopylova L.N., 1988. The geological basis for mineral-technological mapping at
the Kovdor ore deposit. Intern. Geol. Rev., 30, 307-319.
Krasnova N.I., Balaganskaya E.G. and Garcia D., in press. Kovdor – classic phoscorites and
carbonatites. In: Phoscorites and Carbonatites from mantle to mine. (Eds: F Wall and A N
Zaitsev). Mineralogical Society Series 10.
Kukharenko, A.A., 1967. Alkaline magmatism of the Eastern Baltic Shield. Zapiski Vsesoyuznogo
Mineralogicheskogo Obshchestva 4, 393-397. (in Russian).
Kukharenko, A.A., Orlova, M.P., Bulakh, A.G., Bagdasarov, E.A., Rimskaya-Korsakova, O.M.,
Nefedov, E.I., Il’insky, G.A., Sergeev, A.S. and Abakumova, N.B., 1965. The Caledonian
 complex of ultrabasic, alkaline rocks and carbonatites of the Kola Peninsula and northern
Karelia. Moscow, Nedra, 550 p. (in Russian).
Kukharenko, A.A., Bulakh, A.G., Il’insky, G.A., Shinkarev N.F.and Orlova, M.P., 1971.
Metallogenic features of the alkaline formations on the east part of the Baltic shield.
Proseedings of Leningrad natural scientists society. V. LXXII. Iss. 2. 277 p. (in Russian).
Lapin, A.V., 1979. Mineral parageneses of apatite ores and carbonatites of the Seblyavr massif. Int.
Geol. Rev. 21, 1043-1052.
Lee M.-J., Garcia D., Moutte J., Williams C. T. and Wall F., in press. Carbonatites and phoscorites
from the Sokli Complex, Finland. In: Phoscorites and Carbonatites from mantle to mine.
(Eds: F Wall and A N Zaitsev). Mineralogical Society Series 10.
Lee W. J. and Wyllie P. J., 2000. The system CaO-MgO-SiO2-CO2 at 1 GPa, metasomatic
wehrlites, and primary carbonatitic magmas. Contrib. Mineral. Petrol. 138, 214-228.
Liferovich, R.P., 1998. Scandium resources and hydrothermal phosphate mineralization of the
Kovdor carbonatite-phoscorite complex. Extended abstract of Ph.D. dissertation. Moscow.
20 p. (in Russian).
Mahotkin I. L., Downes H., Hegner E. and Beard A., 2003. Devonian dike swarms of alkaline,
carbonatitic and primitive magma-type rocks from the south Kola Peninsula: geochemical
and isotopic constraints on plume-lithospheric interaction. 8th International Kimberlite
Conference Long Abstract. FLA_0346.
Makhotkin I. L., Gibson S. A., Thompson R. N., Zhuralev D. Z. and Zherdev P. U., 2000. Late
Devonian diamondiferous kimberlite and alkaline picrite (?proto-kimberlite) – Magmatism
in the Arkhangelsk region, NW Russia. J. Petrol. 41, 201-227.
Markwick A. J. W. and Downes H., 2000. Lower crustal xenoliths from Arkhangelsk kimberlite
pipes: petrological, geophysical and geochemical constraints from xenoliths. Lithos 51, 135-
151.
Meiboom A., Anderson D.L., Sleep N.H., Frei R., Hren M. T. and Wooden J. L., 2003. Are the high
3
He/4He ratios in oceanic basalts an indicator of deep-mantle plume components? Earth
Planet. Sci. Lett. 208, 197-204.
Nakamura N., 1974. Determination of REE, Ba, Fe, Mg, Na and K in carbonaceous and ordinary
chondrites. Geochim. Cosmochim. Acta. 38, 757-775.
Pyatenko I. K. and Osokin Ye. D, 1988. Geochemical features of the carbonatite palaeovolcano at
Kontozero, Kola Peninsula. Geokhimiya 5, 723-737 (in Russian).
Pilipiuk A. N., Ivanikov V. V. and Bulakh A. G., 2001. Unusual rocks and mineralisation in a new
carbonatite complex at Kandaguba, Kola Peninsula, Russia. Lithos 56, 333-347.
Rimskaja-Korsakova O.M., 1963. About genesis of Kovdor iron-ore deposit. In: Questions of
magmatism and metamorphism. Leningrad: LGU, 1, 125-143. (In Russian).
Rock N M S1991. Lamprophyres. Blackie. Glasgow. 285pp.
Rukhlov A. S., 1999. Dykes and explosive pipes of the Kandalaksha Graben (Kola Alkaline
Province): The models of magmatic processes and evolution of the subcontinental mantle.
PhD thesis. St Petersburg University. 283 pp.
Sablukov, S. M., 1984. Phases of formation and age of diatremes of the Onega Peninsula. Dokl.
Acad. Nauk. 227, 168-170. (in Russian).
Sablukov , S. M., Sablukova L. I. and Shavyrina M. V., 2000. Mantle xenoliths from the Zimnii
Bereg kimberlite deposits of rounded diamonds, Arkhangelsk Diamondiferous Province.
Petrology 8 (5), 466-494.
Schablinsky G.N., 1963. To the deep structure of the Khibiny and Lovozero plutons. In:
Proceedings of Leningrad natural scientists society. V. LXXIV. Iss. 1, 41-43 (in Russian).
Schpachenko A.K. and Stepanov V.A., 1991. A new data on the morphology of the ore-bearing
structures at the north segment of the Khibiny ijolite-urtite arc. Doklady Acad. Nauk., 319.
No 6. 1413-1417. (In Russian).
Sindern S., Zaitsev A. N., Demény A., Bell K., Chakmouradian A. R., Kramm U., Moutte J. and
Rukhlov A. S., 2004. Mineralogy and geochemistry of silicate dyke rocks associated with
 carbonatites from the Khibina complex (Kola, Russia) – isotope constraints on genesis and
small-scale mantle sources. Mineral. and Petrol. 80, 215-239.
Sitnikova M. A., 2004. Mineralogy of carbonatites from the Sallanlatva Massif, Kola Peninsula.
Extended abstract of PhD dissertation. St Petersburg, University of St Petersburg
publications, 16p (in Russian).
Sokolov S.V., 1983. Genetic affinity apatite-magnetite ores and carbonatites of alkaline-ultramafic
massifs. Geokhimiya 3, 438-449.
Subbotin V.V. and Subbotina G.F., 2000. Minerals of pyrochlore group in phoscorites and
carbonatites of Kola Peninsula. Vestnik MGTU. 3 (2), 273-284.
Sun S.-S. and McDonough W. F., 1989. Chemical and isotopic systematics of oceanic basalts:
implications for mantle composition and processes. In: Magmatism in ocean basins. Geol.
Soc Sp. Pub. 42, 313-345.
Ternovoy V.I., Afanasiev B.V. and Sulimov B.I., 1969. Geology and prospecting of the Kovdor
Vermiculite-Phlogopite deposit. Nedra. Leningrad. 288pp. (in Russian).
Thibault Y., Edgar A. D. and Lloyd F. E., 1992. Experimental investigation of melts from a
carbonated phlogopite lherzolite: implications for metasomatism in the continental
lithospheric mantle. American Mineralogist, 77, 784-794.
Tolstikhin I.N. and Marty B., 1998. The evolution of terrestrial volatiles: A view from helium,
neon, argon and nitrogen isotope modelling. Chemical Geol., 147, 27-52.
Tolstikhin, I.N., Kamensky, I.L., Marty, B., Nivin, V.A., Vetrin, V.R., Balaganskaya, E.G., Ikorsky,
S.V., Gannibal, M.A., Kirnarsky, Yu.M., Weiss, D., Verhulst, A. and Demaiffe, D., 1999.
Low mantle plume component in Devonian Kola ultrabasic-alkaline-carbonatite complexes:
Evidences from rare gas isotopes and related parent elements. Apatity-Nancy-Bruxelles.
Printed in CRPG, Nancy. 97 p. (in Russian and English).
Tolstikhin, I.N., Kamensky, I.L., Marty, B., Nivin, V.A., Vetrin, V.R., Balaganskaya, E.G., Ikorsky,
S.V., Gannibal, M.A., Weiss, D., Verhulst, A. and Demaiffe, D., 2002. Gas, rare isotopes
and parent trace elements in ultrabasic-alkaline-carbonatite complexes, Kola Peninsula:
identification of lower mantle plume component. Geochim. Cosmochim. Acta, 66, 881-901.
Torsvik T. H., Smethurst M.A., Meert J. G., Van der Voo, R., McKerrow W. S., Brasier M. D.,
Sturt B. A. and Walderhaug H. J., 1996. Continental break-up and collision in the
Neoproterozoic and Palaeozoic – A tale of Baltica and Laurentia. Earth Science Reviews,
40, 229-258.
Uvad'ev L. I., 1980. Riphean alkaline-magmatism in the zone of the Kandalaksha graben. Doklady
Academii Nauk SSSR 254, 1435-1456.
Vartiainen H. and Woolley A. R., 1974. The age of the Sokli carbonatite, Finland, and some
relationships of the North Atlantic alkaline igneous province. Bull. Geol;. Surv. Finland, 46,
81-91.
Vartiainen H. and Paarma H., 1979. Geological characteristics of the Sokli carbonatite complex,
Finland. Economic Geology, 74, 1296-1306.
Vartiainen H., 1980. The petrography, mineralogy and petrochemistry of the Sokli carbonatite
massif, northern Finland. Bull. Geol. Surv. Finland, 313, 126p.
Veksler I. V., Nielsen T. F. D and Sokolov S. V., 1998. Mineralogy of crystallised melt inclusions
from Gardinier and Kovdor ultramafic alkaline complexes: implications for carbonatite
genesis. J. Petrol. 39, 2015-2031.
Verhulst A., Balaganskaya E., Kirnarsky Yu. and Demaiffe D., 2000. Petrological and geochemical
(trace elements and Sr-Nd isotopes) characteristics of the Paleozoic Kovdor ultramafic,
alkaline and carbonatite intrusion (Kola Peninsula, NW Russia). Lithos 51, 1-25.
Wall F., Zaitsev A.N. and Mariano A.N, 2001. Rare earth pegmatites in carbonatites. J. African
Earth Sci., 32, 35-36.
Wall F. and Zaitsev A. N. (in press). Rare earth minerals in Kola carbonatites. In: Phoscorites and
Carbonatites from mantle to mine. (Eds: F Wall and A N Zaitsev). Mineralogical Society
Series 10.
Wallace M. E. and Green D. H., 1988. An experimental determination of primary carbonatite
magma composition. Nature 335, 343-346.
Woolley, A.R. 1989. The spatial and temporal distribution of carbonatites. In: Bell, K. (ed.).
Carbonatites: Genesis and Evolution. Unwin Hyman, London, 15-37.
Wyllie, P.J., Baker, M.B. and White, B.S., 1990. Experimental boundaries for the origin and
evolution of carbonatites. Lithos, 26, 3-19.
Zaitsev, A. and Bell K., 1995. Sr and Nd isotope data of apatite, calcite and dolomite as indicators
of source, and the relationships of phoscorites and carbonatites from the Kovdor massif,
Kola peninsula, Russia. Contrib. Mineral. Petrol., 121, 324-335.
Zaitsev A.N., 1996. Rhombohedral carbonates from carbonatites of the Khibiny massif, Kola
Peninsula, Russia. Can.Miner., 34, 453-468.
Zaitsev, A.N., Bell, K., Wall, F. and Le Bas, M.J., 1997. Alkaline-rare earth carbonates from
carbonatites of Khibiny massif: Mineralogy and genesis. Transactions (Doklady) Russian
Akad. Sci./Earth Sci. Sect. 355 (5), 786-790.
Zaitsev, A.N., Sitnikova M. A., Subbotin V. V., Fernandes-Suarez J. and Jeffries T. E., in press.
Sallanlatvi complex – a rare example of magnesite and siderite carbonatites. In: Phoscorites
and Carbonatites from mantle to mine. (Eds: F Wall and A N Zaitsev). Mineralogical
Society Series 10.