mates of the closure temperature ranging from 600 Mezger et al., 1992 to 750°C Zhou and
Hensen, 1995. Daly et al. in review report SmNd ages for different petrographic varieties of
garnet whose PT history has been inferred from reaction textures and determined independently
using conventional thermobarometry. In one rock, the garnet SmNd age is identical to that
for UPb in metamorphic zircon and 20 Ma older than a concordant UPb monazite age, indicating
a high closure temperature, above about 650 – 700°C for the garnet SmNd system. In this case
the SmNd system seems to be dating metamor- phic events that correlate with the petrography.
This is consistent with some studies, such as those of Vance and O’Nions 1990 and Hensen and
Zhou 1995, which have concluded that the SmNd system is capable of dating garnet crys-
tallisation during high grade metamorphism at temperatures up to 700°C and of surviving net
transfer reactions in the same rock at tempera- tures as high as 500°C. However these studies
disagree with the conclusion of Mezger et al. 1992 that the garnet SmNd system was limited
by closure temperature, which they argued must be as low as 600°C to be consistent with their data
from the Adirondacks. SmNd mineral isotopic data Table 2 are
presented for both the leucosome and mesosome from one migmatite sample 21, Fig. 3. Garnet in
this rock exhibits only one petrographic variety. The leucosome sample 21A yields a garnet-
whole-rock SmNd age of 187096.5 Ma. Gar- nets from the mesosome sample 21R yield an
identical SmNd age garnet–WR of 187096.4 Ma. Combining these data, and including analy-
ses of a feldspar separate, yields a combined SmNd isochron age of 187097 Ma MSWD=
2.0. These data provided a minimum age for garnet growth during melting and M2 metamor-
phism see above. The M1 metamorphism has not been dated in this study and remains to be
evaluated.
3. Discussion
SmNd model ages presented here for the Lap- land Granulite Terrane are similar to those re-
ported by Huhma and Merila¨inen 1991 and to a t
DM
age of 2.55 Ga Huhma, 1986 for the post- tectonic 1.77 Ga Nattanen granite close to the
southern border of the Lapland Granulite Ter- rane. Based on c 2.3 Ga model ages for metasedi-
Fig. 11. Schematic cross sections from the Kolvitsa Belt to the UGT Fig. 4 and from the UGT to the Imandra-Varzuga Belt Fig. 2. KB, Kolvitsa Belt; LKS, Lapland – Kola Suture; PS, Peschanoozerskaya Suite; SU, Sergozerskaya Unit. PT data and inverted
metamorphic gradient from Timmerman 1996 and unpublished data and Belyayev et al. 1977.
mentary samples, Huhma and Merila¨inen 1991 also concluded that the protoliths of the Lapland
Granulite Terrane were predominantly Palaeo- proterozoic in age. The SmNd data are also in
agreement with previous inferences based on UPb analyses of detrital zircon which yielded
207
Pb
206
Pb and UPb ages between 2.0 and 2.15 Ga for samples containing no Archaean zircons
Merila¨inen, 1976; Sorjonen-Ward et al., 1994 and between 2.0 and 3.6 Ga for samples with
zircons from Archaean sources Tuisku and Huhma, 1998a; Bridgwater et al., 1999. Palaeo-
proterozoic t
DM
ages ranging from 2.45 to 2.13 Ga Daly et al., 1997 also characterise the
metasediments of the UGT, generally regarded as a southeastwards extension correlative of the
Lapland Granulite Terrane Fig. 1 as well as the Tersk Terrane as shown above.
Thus, Palaeoproterozoic
metasedimentary rocks within the core zone of the LKO have
positive to weakly negative initial o
Nd
values over a strike length of at least 600 km. Adjacent Ar-
chaean terranes — including the Murmansk, Central Kola, and Belomorian terranes Balagan-
sky et al., 1998a all have more strongly negative
o
Nd
signatures Fig. 7, Timmerman and Daly, 1995. This shows that although Archaean detrital
zircons are present Tuisku and Huhma, 1998a; Bridgwater et al., 1999 the surrounding Archaean
regions have
contributed only
subordinate amounts of detritus to the metasediments of the
Lapland Granulite, Umba Granulite and Tersk terranes. Importantly, metaigneous rocks within
these terranes, dated at c 1.96 Ga in the Tersk Terrane this paper and less precisely between
1.90 and 1.93 Ga in the Lapland Granulite Ter- rane Sorjonen-Ward et al., 1994 and at 1.91 –
1.94 Ga in the UGT Umba Complex, Kislitsyn et al., 1999b; Glebovitsky et al., 2000 also exhibit
positive to weakly negative o
Nd
values Fig. 8. These data clearly demonstrate the presence of
large volumes of juvenile Palaeoproterozoic crust within the core zone of the orogen. As previously
suggested by Barbey et al. 1984, based on geo- chemical evidence, we draw the obvious conclu-
sion from these results that the juvenile protoliths of the Lapland Granulite, Umba and Tersk ter-
ranes developed as the products of arc magma- tism,
as a
result of
subduction of
the ‘Lapland – Kola’ ocean, which itself originated by
oceanic separation following rifting and terrane dispersal initiated c 2.45 Ga ago.
This model accounts for the relative lack of Archaean detritus in the Lapland – Kola metasedi-
ments as well as for the arc-signature of calc-alka- line magmatism in the Lapland Granulite Barbey
and Raith, 1990, Inari Barling et al., 1997 and Tersk Ivanov, 1987; Daly and Brewer, unpub-
lished data terranes. Subduction polarity in the western Kola Peninsula was probably northward-
directed as previously suggested by Barbey et al. 1984 and substantiated by geochemical and
geochronological investigations of calc-alkaline magmatism within the Inari terrane in the hang-
ing wall of the LKO Barling et al., 1997, dated at c 1.94 – 1.91 Ga Barling et al., 1997; Tuisku
and Huhma 1998a.
Subsequent collision has preserved the footwall Belomorian Terrane, parts of the rifted margin
Tanaelv Belt, arc and possibly both fore-arc and back-arc sedimentary basins Lapland Granulite
Terrane as well as Andean-margin subduction-re- lated magmatism in the Inari Terrane. The timing
of the collisional event within the Lapland Gran- ulite Terrane requires refinement but granulite-fa-
cies metamorphic zircons suggest deep burial by c 1.9 Ga e.g. Sorjonen-Ward et al., 1994 and
decompressional melting at or before c 1.87 Ga, as discussed above.
Shallow seismic reflection data across the Lap- land Granulite Terrane reveal strong north-dip-
ping reflectors parallel to near-surface tectonic structures and lithological layering Korja et al.,
1996. Refraction data, e.g. from the POLAR profile, have been interpreted to show that the
Lapland Granulite Terrane is a superficial struc- ture consistent with gravity modelling. However,
lateral variations in deep crustal seismic velocity and VpVs ratio Walther and Fleuh, 1993 to-
gether with reflections traversing the entire crust revealed by reprocessing the Polar Profile data
Pilipenko et al., 1999, suggest the presence of a major trans-crustal structure implying that the
Lapland Granulite Terrane extends to mantle depths. We suggest that this structure — the
Lapland – Kola Suture LKS — represents the
suture zone of the LKO, possibly connected at depth to a fossilized subduction zone still pre-
served within the mantle. As shown in Fig. 1, we place the footwall boundary of the LKS between
the Lapland Granulite Terrane and the Tanaelv Belt.
To the east the correlative units within the core zone of the LKO — the Umba and Tersk Ter-
ranes and the Kolvitsa Belt — display a more complex structure, aspects of which are discussed
in detail elsewhere e.g. Timmerman, 1996; Bala- gansky et al. 1998a,b, Balagansky et al. 2000.
Within much of the region between the southeast- ern end of the LGT and the Kolvitsa Belt Fig. 1,
the dominant Palaeoproterozoic structure is sub- horizontal.
Moving eastwards
through the
Kolvitsa Belt into the UGT, lineations plunge east – southeastwards while further east still, in the
Tersk Terrane Figs. 1, 2 and 11 the major structures dip southwards. This complex struc-
tural pattern Figs. 1 and 11 probably takes the form of a large-scale compressional flower struc-
ture, modified by later extensional deformation. Our
own structural
observations along
the Varzuga River section Fig. 2 and interpretation
of potential field data Balagansky et al. 1998a suggest that subduction polarity in the central and
eastern part of the Kola peninsula may have been southwards.
Along the Varzuga River section Fig. 2, or- thogneisses of the Strelna Domain that occur
south of the Imandra – Varzuga Belt are Archaean in age as confirmed by the new UPb zircon ages
for samples 895-59 and 895-80. Broadly similar results were obtained by Balashov et al. 1992
who reported a UPb zircon age of 2670910 Ma and a RbSr whole rock isochron age of 28709
29 Ma for granitic gneisses from the Babya River area further east in the Strel’na Domain. Confir-
mation of the Neoarchaean age for the TTG gneisses of the Strel’na domain and the likely
Palaeoproterozoic depositional age for the Ser- gozerskaya sediments of the Tersk Terrane sup-
ports the conclusion of Balagansky et al. 1998a that a major tectonic boundary exists between
them Fig. 2. This boundary has not been mapped in detail. However, it coincides with a
major break in both the gravity and magnetic signatures. Close to the Varzuga River Fig. 2,
this potential field feature dips south – southwest- wards Mints et al., 1996 and has a west – north-
westerly trend parallel to the strike of foliation and
lithological layering.
Further east,
the boundary swings into a northwest – southeast ori-
entation Balagansky et al., 1998a. Structural observations are consistent with reverse thrust
motion along this boundary.
Southward-directed subduction in the central and eastern part of the Kola Peninsula is consis-
tent with all structural observations Figs. 2 and 11, Fedorov et al., 1980. A southwards dip for
the main fault underlying the Umba Granulite Terrane within the suture zone Fig. 4 was previ-
ously suggested by Glaznev et al. 1997. The change in subduction polarity and in the dip of
the LKS from southwards in the east beneath the Tersk Terrane to northwards in the west beneath
the Lapland Granulite Terrane may reflect an original offset in the rifted margin of the Lap-
land – Kola ocean. The position of this proposed offset corresponds to the location of the rift-re-
lated c 2.45 Ga Main Ridge massif of gabbro anorthosite Fig. 1, which was emplaced into a
releasing bend during dextral transtension Bala- gansky et al., 1998a.
The main deformation and amphibolite-facies migmatisation Belyayev et al. 1977 in the
Varzuga River region is bracketed by 1.96 Ga, the age of arc magmatism in the Tersk Terrane, and
1.90 – 1.92 Ga, the age of late-tectonic pegmatites, the more reliable of which has an age of 1907 9
10 Ma. Available UPb geochronology suggests that deformation and metamorphism took place
within a similar time interval throughout the Strelna – Tersk – Umba – Kolvitsa region, though
the grade of metamorphism varies considerably. For example, a high-grade leucosome that post-
dates the high-pressure granulite-facies metamor- phism within the collisional me´lange between the
Umba Terrane and the Kolvitsa Belt Kislitsyn et al., 1999a has also yielded a 1.91-Ga age. Meta-
morphic zircon ages of 1.90 – 1.92 Ga from the Kochinny Cape area Frisch et al., 1995; Kaulina,
1996, further to the west in the footwall of the LKS, reflect this collision but under high-pressure
and high-temperature amphibolite-facies condi- tions Alexejev, 1997.
Field, petrographic and thermobarometric evi- dence Fig. 11, Krill, 1985; Barbey and Raith,
1990; Timmerman, 1996 and unpublished data demonstrate that inverted metamorphic gradients
characterise the suture zone of the LKO along its entire length from Norway to the White Sea.
Evaluation of this phenomenon is beyond the scope of this paper but we note that inverted
metamorphism is also well documented in several major collisional orogens and suture zones rang-
ing from Cenozoic to Palaeoproterozoic in age, e.g. Himalayas Searle and Rex, 1989, Variscan
Belt Burg et al., 1989, Grenville Belt Brown et al., 1992 and in the Cheyenne Belt of the south-
western USA Duebendorfer, 1988.
4. Conclusions
