Mesoarchean Kola-Karelia continent

Downloaded from specialpapers.gsapubs.org on July 8, 2015

CHAPTER 2

Mesoarchean Kola-Karelia continent Michael V. Mints Geological Institute, Russian Academy of Sciences (RAS), 7 Pyzhevsky Lane, Moscow, 119017, Russia Ksenia A. Dokukina Geological Institute, Russian Academy of Sciences (RAS), 7 Pyzhevsky Lane, Moscow, 119017, Russia, and Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow, 119991, Russia Alexander N. Konilov Geological Institute, Russian Academy of Sciences (RAS), 7 Pyzhevsky Lane, Moscow, 119017, Russia, and Institute of Experimental Mineralogy, Russian Academy of Sciences (RAS), Chernogolovka, Moscow Region, 142432, Russia Tatiana V. Kaulina Geological Institute, Kola Science Centre RAS, 14 Fersman Str., Apatity, Murmansk Region, 184209, Russia Elena A. Belousova GEMOC ARC National Key Centre, Department of Earth and Planetary Sciences, Macquarie University, Sydney, NSW 2109, Australia Peter A. Dokukin Peoples’ Friendship University of Russia, 6 Miklukho-Maklaya Str., Moscow, 117198, Russia Lev M. Natapov GEMOC ARC National Key Centre, Department of Earth and Planetary Sciences, Macquarie University, Sydney, NSW 2109, Australia Konstantin V. Van Institute of Experimental Mineralogy, Russian Academy of Sciences (RAS), Chernogolovka, Moscow Region, 142432, Russia

ABSTRACT The Mesoarchean Kola-Karelia continent in the eastern Fennoscandian Shield includes three tectonic provinces, Kola, Karelia and Belomoria, that were formed by the Paleoarchean and Mesoarchean microcontinents. Traces of Mesoarchean tonalite-trondhjemite-granodiorite (TTG)−type early crust were documented in all of the most ancient units of the Kola-Karelia continent. Ancient crust was revealed and dated in the Ranua and Iisalmi microcontinents, 3.5–3.4 Ga;

Mints, M.V., Dokukina, K.A., Konilov, A.N., Kaulina, T.V., Belousova, E.A., Dokukin, P.A., Natapov, L.M., and Van, K.V., 2015, Mesoarchean Kola-Karelia continent, in Mints, M.V., Dokukina, K.A., Konilov, A.N., Philippova, I.B., Zlobin, V.L., Babayants, P.S., Belousova, E.A., Blokh, Y.I., Bogina, M.M., Bush, W.A., Dokukin, P.A., Kaulina, T.V., Natapov, L.M., Piip, V.B., Stupak, V.M., Suleimanov, A.K., Trusov, A.A., Van, K.V., and Zamozhniaya, N.G., East European Craton: Early Precambrian History and 3D Models of Deep Crustal Structure: Geological Society of America Special Paper 510, p. 15–88, doi:10.1130/2015.2510(02). For permission to copy, contact [email protected]. © 2015 The Geological Society of America. All rights reserved.

15


16

Chapter 2 Vodlozero and Khetolambina microcontinents, 3.25–3.15 Ga; Kuhmo-Segozero microcontinent, ~3.0 Ga; Murmansk and Inari-Kola microcontinents, 2.93 Ga; and Kianta microcontinent, 2.83–2.81 Ga. In the older (>3.0 Ga) tectonic units and microcontinents, the ancient crust was possibly formed in brief bursts of endogenic activity. In younger microcontinents (3.0–2.93 Ga), these processes could continue until 2.8 and even 2.72 Ga. The tectonic settings in which early TTG crust has been produced are largely uncertain. The primary melt glassy inclusions with a glass phase in cores of prismatic zircon crystals from TTG gneisses provide evidence for the volcanic origin of gneiss protolith. Suggested genetic modeling of TTG-type complexes assumes that felsic K-Na melts with positive Eu anomaly are a product of dry high-temperature partial melting of the previously formed mafic-to-felsic crustal rocks and/or thick older TTG crust. Positive Eu anomaly in the eutectic is directly related to the predominance of plagioclase and K-feldspar in the melt. TTG-type crust melted to produce granitegranodiorite (GG) rocks. Earliest microcontinents are separated by Mesoarchean greenstone belts (mainly 3.05–2.85 Ga, in some cases up to 2.75 Ga), which are fragments of paleo–islandarc systems accreted to their margins: the Kolmozero-Voronya, Central Belomorian, Vedlozero-Segozero, Sumozero-Kenozero, and Tipasjärvi-Kuhmo-Suomussalmi belts; and the mature island arcs (microcontinents): Khetolambina and Kovdozero. These structural units are characterized by significant extent, close to rectilinear trend, localization along the boundaries between Archean microcontinents, and a specific set of petrotectonic assemblages (basalt-andesite-rhyolite, komatiite-tholeiite, and andesite-dacite associations). The recently discovered Meso-Neoarchean Belomorian eclogite province that is structurally linked with the Central Belomorian greenstone belt contains two eclogite associations distributed within TTG gneisses: the subduction-type Salma association and the Gridino eclogitized mafic dikes. The protolith of the Salma eclogites is thought to have been a sequence of gabbro, Fe-Ti gabbro, and troctolite, formed at ca. 2.9 Ga in a slow-spreading ridge (similar to the Southwest Indian Ridge). The main subduction and eclogite-facies events occurred between ca. 2.87 and ca. 2.82 Ga. Mafic magma injections into the crust of the active margin that led to formation of the Grigino dike swarm were associated with emplacement of a mid-ocean ridge in a subduction zone, beginning at ca. 2.87 Ga. Crustal delamination of the active margin and subsequent involvement of the lower crust in subduction 2.87–2.82 Ga ago led to high-pressure metamorphism of the Gridino dikes that reached eclogite-facies conditions during a collision event between 2.82 and 2.78 Ga. This collision resulted in consolidation of the Karelia, Kola, and Khetolamba blocks and formation of the Mesoarchean Belomorian accretionary-collisional orogen. To date, the subductionrelated Salma eclogites provide the most complete and meaningful information on the nature of plate tectonics in the Archean, from ocean-floor spreading to subduction and collision. The Kovdozero granite-greenstone terrain that separates the Khetolambina and Kuhmo-Segozero microcontinents is formed by TTG granitoids and gneisses hosting metasediments and metavolcanics of several greenstone belts, which belonged to the Parandovo-Tiksheozero island arc that existed from ca. 2.81 to 2.77 Ga. The Iringora greenstone belt includes the ophiolite complex of the same name with an age of 2.78 Ga. The collision of microcontinents resulted in the upward squeezing of the island arc and the obduction of its marginal portions onto surrounding structures.


Mesoarchean Kola-Karelia continent

17

2. INTRODUCTION Michael V. Mints

The complexly built Archean Kola-Karelia continent occupies the northern part of the East European craton (Fig. 0.2A). The results of geological mapping and tectonic and metallogenic analyses are summarized in regional maps (Smyslov, 1978; Popov and Rundqvist, 1984; Silvennoinen et al., 1987; Koistinen et al., 2001; Glebovitsky, 1991) and published in monographs and papers that characterize the tectonics, deep structure, and evolution of the eastern Fennoscandian Shield (Kratz et al., 1978; Sidorenko and Bilibina, 1980; Popov, 1986; Sokolov, 1987; Rundquist and Mitrofanov, 1993; Mitrofanov et al., 1995b; Mints et al., 1996; Sharov, 1997; Mitrofanov and Sharov, 1998; Ermolaeva, 2001; Glaznev, 2003; Glebovitsky, 2005). The Archean complexes in the western part of the region have been described by Finnish and Norwegian colleagues (Bjørlykke et al., 1985; Hölttä et al., 2000, 2014; Sorjonen-Ward and Luukkonen, 2005; Saltikoff et al., 2006; Kukkonen and Lahtinen, 2006). The Archean crust of the eastern Fennoscandian Shield is conventionally regarded as a combination of three tectonic provinces: the Belomorian mobile belt, sandwiched between the Kola and Karelian stable blocks (cratons; Kratz et al., 1978; Rundquist and Mitrofanov, 1993; Mitrofanov et al., 1995b; Glebovitsky, 2005; Slabunov et al., 2006b, 2006d; Hölttä et al., 2000, 2014; Sorjonen-Ward and Luukkonen, 2005; Mints et al., 2010b). Further subdivision and demarcation of the Archean crust within the Kola-Karelia continent are complex tasks, because boundaries between rocks have been repeatedly reworked (deformation, metamorphism, magmatic activity) and overlapped by tectonic nappes and volcanic-sedimentary complexes throughout the Archean and Paleoproterozoic. Multiple complications of the crustal structure indicate that the boundaries of tectonic units inherit older borders or conversely crosscut ancient crustal domains and boundaries between them. Until recently, researchers of the eastern Fennoscandian Shield, postulating a plate-tectonic approach to reconstruction of early Precambrian geological history, have proposed various versions of territorial (in fact, geographic) subdivision of the early Precambrian crust of the Kola-Karelia continent into fault-bounded areas (provinces, subprovinces, terranes, blocks, etc.), in which the geological history differs from that of the adjacent areas. The division is mostly based on lithology and age of crustal rock associations. It should be noted that, except for rare cases, the inferred boundaries of Archean crustal domains shown in geological and tectonic maps or hidden beneath Paleoproterozoic complexes remain conventional. The recognition of the main stages in the evolution of the Archean crust of Kola-Karelia continent is closely related to the study of temporal, geodynamic, and evolutionary relationships between complexes of low- and medium-grade metamor-

phic rocks, on the one hand, and high-grade rocks, on the other hand; i.e., between granite-greenstone terranes and granulitegneiss belts in a broad sense of these terms. Our recent studies in the eastern Fennoscandian Shield (Mints et al., 2010b) have convincingly shown that manifestations of Archean granulitefacies metamorphism and attendant high-temperature magmatism in intrusive and volcanic facies are separated in time from the period when tonalite-trondhjemite-granodiorite (TTG)–type crust was growing. The metamorphic and magmatic processes developed within a continental domain made up of microcontinents that eventually amalgamated into the Kola-Karelia continent during the final period of Mesoarchean crustal growth. The onset of this period varies in particular domains of the region near 2.75 Ga (Lobach-Zhuchenko and Chekulaev, 2007; Mints et al., 2010b). Two groups of evidence make it possible to identify and delineate the crustal domains in a tectonic map that corresponds to the Paleoarchean–Mesoarchean stage of origin and growth of the continental crust. The first group includes isotopic geochronological data, which point out the crustal blocks composed of pregreenstone TTG gneisses formed over short time intervals between ca. 3.5 and ca. 2.8 Ga or indicate participation of rocks having that age as a source of partial melts and paleosome of migmatites (Lobach-Zhuchenko et al., 1989; Levchenkov et al., 1995; Sergeev et al., 2007a, 2008; Chekulaev et al., 2009; Kröner and Compston, 1990; Jahn et al., 1984; Juopperi and Vaasjoki, 2001; Sorjonen-Ward and Luukkonen, 2005; Peltonen et al., 2006; Mints et al., 2010b; Hölttä et al., 2000, 2012a, 2012b, 2014). It is known that Archean greenstone belts in the KolaKarelia are variable in age, lithology, and structure. Various inter- and intracontinental geodynamic settings are suggested for particular belts (Lobach-Zhuchenko, 1988; Lobach-Zhuchenko et al., 1993b, 1999, 2000a; Luukkonen, 1988; Kulikov et al., 1990; Svetov, 1997, 2005; Papunen et al., 1998; Puchtel et al., 1998a, 1999, 2001; Samsonov et al., 1996, 2005; Kudryashov et al., 1999; Kozhevnikov, 2000; Matrenichev et al., 2000; Svetov et al., 2001; Bibikova et al., 2003; Chashchin et al., 2004; Kozhevnikov et al., 2006; Slabunov et al., 2006a, 2006d, 2007; Mil’kevich et al., 2007; Shchipansky, 2008; Shchipansky et al., 2004; Huhma et al., 2012b). General trends in arrangement of Archean greenstone belts by their age and role in the geological history of the Kola-Karelia continent remain poorly studied. As a result of joint consideration of geological, geochemical, and geochronological data, Mints et al. (2010b) recognized two types of greenstone belts: (1) the Mesoarchean belts, which are formed by fragments of island-arc systems accreted to the margins of ancient


18

Chapter 2

microcontinents (continental nuclei) and the mature island arcs (these structural units are characterized by significant extent, close to rectilinear trend, localization along the boundaries between Archean microcontinents, and by a specific set of rocks); and (2) the Neoarchean greenstone belts, mainly related to epicontinental rifting. Thus, the Mesoarchean belts make up the second group of evidence that allows us to distinguish the crustal domains, which include relics of the Paleoarchean and Mesoarchean embryos of microcontinents. Finally, the results presented in this book make it possible to recognize the main constituents of the Archean crust in eastern Fennoscandia, corresponding to three periods of its evolution. Their chronological boundaries are somewhat variable over a vast region. These are (1) Paleoarchean and Mesoarchean embryos of microcontinents (pregreenstone, mostly TTG crust); (2) Mesoarchean greenstone belts composed of island-arc and marginal continental complexes with fragments of the oceanic crust in form of ophiolites and eclogites; and (3) Neoarchean intracontinental areas of sedimentation, high-temperature metamorphism, and magmatism. The composition, structure, and evolution of those Archean crustal domains, which are related to Paleoarchean and Mesoarchean stages of geological history, are considered here in chapter 2. The geological complexes and tec-

tonic structural units pertaining to the Neoarchean period of tectonothermal activity and localized within the boundaries of the Kola and Karelian-Belomorian areas are presented in chapter 3. The Kola-Karelia continent is an assembly of three groups of Paleoarchean and Mesoarchean microcontinents that take part in the structure of the following provinces (Appendix I-41): (1) Kola Province (Murmansk and Inari-Kola microcontinents); (2) Karelian Province (Vodlozero, Ranua, Iisalmi, Kianta, and Kuhmo-Segozero microcontinents); and (3) Belomorian Province (Khetolambina and Kovdozero microcontinents. Microcontinents are separated from one another by early greenstone belts, which are interpreted as accretionary orogens and sutures: Kolmozero-Voronya, Central Belomorian, Vedlozero-Segozero, Sumozero-Kenozero, and TipasjärviKuhmo-Suomussalmi belts, as well as the Khetolambina and Kovdozero mature island arcs. We have to emphasize, once again, that our tectonic subdivision of the Paleoarchean and Mesoarchean crust differs in the sense and form from earlier proposed variants for both the Kola Peninsula (Radchenko et al., 1992; Balagansky et al., 1998; Mitrofanov, 2001; Glebovitsky, 2005) and Karelia (Slabunov et al., 2006b, 2006d; Sorjonen-Ward and Luukkonen, 2005; Hölttä et al., 2012a, 2012b, 2014).

1

Appendix I (I-1–I-23) is available in reduced format at the end of this volume, on the CD-ROM accompanying this volume, and as GSA Data Repository Item 2015089, which is available at www.geosociety.org/pubs/ft2015.htm, or on request from [email protected] or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, USA.



19

2.1. PALEOARCHEAN AND MESOARCHEAN MICROCONTINENTS Michael V. Mints

2.1.1. Murmansk Microcontinent The Murmansk microcontinent (Fig. 0.2A; Appendix I-1 [on CD-ROM accompanying this volume]) is exposed along the Barents Sea shore of the Kola Peninsula as a block of Archean crust that extends for ~600 km in the northeastern direction and is 60–70 km in width. According to the regional geophysical data, the microcontinent extends offshore beneath sedimentary

cover for ~50 km, and its total width can reach 110–120 km. The term “Murmansk block” is widely used in the literature. The predominance of granitic rocks and migmatites is also reflected in the frequently used terms “Murmansk granitoid massif” (Mints et al., 1996, and references therein) and “Murmansk granitemigmatite complex.” The Murmansk microcontinent (Fig. 2.1) is characterized by the prevalence of diverse migmatites and granitoids partly

Figure 2.1. Tectonic structure of the Neoarchean Murmansk granite-migmatite complex. (1) Kolovai and Ponoy plutons: granite with microcline porphyroblasts, 2.62–2.60 Ga; (2–7) Murmansk microcontinent: (2) Kachalovka greenstone belt: migmatized gneisses and amphibolites, 2.74–2.72 Ga; (3) Iokanga and Port-Arthur plutons: lepidomelane-ferrohastingsite granite and granosyenite, 2.80 Ga; (4–6) Murmansk granite-migmatite complex: (4) Iokanga-type (a) granites and (b) migmatites, 2.77–2.72 Ga; (5) Tireberka-type (a) granites and (b) migmatites with relict granulite-facies mineral assemblages, 2.83–2.70 Ga; (6) areas with skialiths and xenoliths; intrusions of mafic granulites, enderbites, and charnockites, ca. 2.83 Ga; (7) Patchemvarek and Severny gabbro-anorthosite massifs, 2.93 Ga; (8) suture (Kolmozero-Voronya greenstone belt): (a) amphibolite-gneiss-schist complex and (b) migmatites with inclusions of the aforementioned rocks, 2.87–2.73 Ga; (9) master faults: (a) normal and reverse–strike-slip faults and (b) arcuate and semiring reverse and strikeslip faults.


20

Chapter 2

transformed into granite gneisses. The geochemistry of this complex is broadly similar to the TTG series; however, a significant portion of the complex is composed of K-enriched granodioritegranite (GG). The southern boundary is marked by the KolmozeroVoronya greenstone belt (suture) and plunges in the northeastern direction at angles of 40°–80°, being steeper in the central part and more gentle in the southeastern part. The Murmansk microcontinent thrusts southwestward. A limited abundance of the greenstone associations proper does not allow the Murmansk microcontinent to be defined as a granite-greenstone domain. Small gabbro-anorthosite bodies (Mount Patchemvarek is the largest one) localized along the boundary with the KolmozeroVoronya belt also belong to the Murmansk microcontinent. Minor lepidomelane-ferrohastingsite granite intrusions are situated in the eastern Murmansk microcontinent. Minor gabbro, gabbronorite, peridotite, and pyroxenite intrusions, including dikes of various age, have been identified throughout the Murmansk microcontinent (some of them are possibly Neoarchean). Patchemvarek Gabbro-Anorthosite Massif The Patchemvarek gabbro-anorthosite massif, located in the marginal part of the microcontinent, is the oldest. The massif consists of two isolated bodies: the massif of Mount Patchemvarek proper, and the Northern massif, ~5 km to the northeast. The massif of Mount Patchemvarek extends for ~15 km in the northwestern direction. Both massifs are hosted in granite gneisses and migmatites of the Murmansk complex. The boundary between the Patchemvarek massif and the Kolmozero-Voronya belt is tectonic, with evidence for thrusting of the Kolmozero-Voronya complex over the Murmansk microcontinent, including both gabbro-anorthosite massifs. Gabbro-anorthosites of the Mount Patchemvarek massif and the Northern massif are characterized by low rare earth element (REE) concentrations and flat REE patterns with a posi-

No. 1

2 3 4 5 6 7

8

tive Eu anomaly (Eu/Eu* = 1.97–2.24). For gabbro-anorthosite of the Northern massif, initial εNd is +2.65 and 87Sr/86Sr(i) = 0.70102 ± 8. These parameters indicate the juvenile nature of magma and a depleted mantle source (Kudryashov and Mokrushin, 2011). Gabbro-anorthosite of the Patchemvarek massif dated by Kudryashov at 2.93 Ga (Table 2.1) is a fragment of the ancient protolith of pregranulitic and pregranitic associations of the Murmansk microcontinent. Granite-Migmatite Murmansk Complex In the granite-migmatite Murmansk complex (Fig. 2.1; Appendices I-1 and I-4), the substantially sodic (TTG) and potassic-sodic (GG) migmatites and granitic rocks are distinguished. TTG migmatites and granitoids are ubiquitous and the most characteristic of the zone immediately adjoining the longitudinal southwestern boundary of the Murmansk massif. In the rest of the territory, they occupy relatively small relict areas among the GG migmatites. The largest body of orthopyroxene tonalite and enderbite is known near Lake Kanentjavr to the southeast of Kola Bay, and minor two-pyroxene–plagioclase granulite and enderbite bodies are located close to the southwestern boundary of the Murmansk massif. The microclineplagioclase migmatites and granitic rocks are, in turn, controlled by oval and semi-oval structures bounded by arcuate and semicircle centroclinal faults. Granite bodies proper mainly are crescent in plan view and enclosed into arcuate faults or localized in the central parts of ovals (Mints et al., 1996). By analogy with the Volgo-Uralia microcontinent (see chapter 6), it can be suggested that the ovals, ~10–60–80 km in diameter, are bowlshaped synforms. Within large ovals and beyond them, small ovals, 5–20 km across, are localized chaotically. Periclinal, vertical, and centroclinal orientations of gneissic banding are observed along their boundaries, indicating that at least some ovals are domes.

TABLE 2.1. AGE DETERMINATIONS ON ROCKS IN THE MURMANSK MICROCONTINENT Complex, group Rock, process Age (Ga) Method Reference Patchemvarek massif Gabbro-anorthosite (a fragment U-Pb thermal ionization Kudryashov and 2.925 of the ancient protolith of the mass spectrometry Mokrushin (2011) granite-migmatite Murmansk (TIMS) zircon age complex) Kanentjavr massif Orthopyroxene tonalite and U-Pb TIMS zircon age Pushkarev et al. (1978) 2.83–2.82 enderbite Enderbite 2.772 Kozlov et al. (2006) Teriberka-type Granite and pegmatite cutting 2.83–2.81 Pb-Pb and Th-Pb Pushkarev et al. (1978) granitoids, central through gabbro-anorthosite of whole-rock isochron age Murmansk massif the Patchemvarek massif Teriberka-type Granite 2.79 U-Pb TIMS zircon age Petrovsky et al. (2008) granitoids, eastern Murmansk massif Iokanga-type Trondhjemite and diorite 2.77–2.72 Kozlov et al. (2006) granitoids, eastern Murmansk massif Iokanga-type Late granodiorite and pegmatite 2.72–2.62 Nitkina and Kaulina granitoids, western (2001) Murmansk massif Kolovai and Ust-Ponoi massifs

Granitoids

2.62–2.60

U-Pb isochron age (zircon and titanite)

Pushkarev et al. (1978)


Mesoarchean Kola-Karelia continent TTG and GG granitoids form relatively large bodies, mainly with diffuse boundaries, as well as neosome of migmatites in the central Murmansk microcontinent. These are coarse-grained rocks, commonly with blue (due to the finest rutile inclusions) and less frequently light-gray quartz, and with frequently observed relics of granulite-facies mineral associations: orthopyroxene or clinopyroxene, brown hornblende, and red-brown high-Ti biotite. These Tireberka-type (after the Teriberka Gulf) granitic rocks have porphyritic structure (porphyroblastic in the mesosome of migmatites), which is determined by unevenly distributed microcline crystals and partly relict orthoclase up to 1–3 cm in size. The Tireberka-type migmatites and granitic rocks contain xenoliths or skialiths of amphibolized two-pyroxene–plagioclase crystal schists, enderbite, and charnockite. The Iokanga-type (after Iokanga River) fine- to medium-grained inhomogeneous trondhjemite and granite occur in the eastern part of microcontinent in some segments of the Barents Sea coast. These granitoids make up bodies with diffuse boundaries or neosome of migmatites. Sporadic blue quartz typical of postgranulite granites is noted. Xenoliths or skialiths of amphibolite are abundant in the Iokanga-type granite. According to Mints et al. (1996), granitoids and migmatites of the Murmansk complex are variable in composition from tonalite to K-Na– and K-leucogranite (64.2–75.7 wt% SiO2, 5.2–8.9 wt% Na2O + K2O); Na2O/K2O = 3.3–0.63. REEs are characterized by sharply differentiated patterns with (La/Yb)N = 40–215 and low heavy (H) REE contents (YbN = 5–7 to 0.4–0.6). Most patterns reveal a positive Eu anomaly (Eu/Eu* = 0.8–3.3, occasionally up to 5.0). The highest Eu anomaly is related to the high-silica granites, which are depleted in both HREEs and light (L) REEs. A typical example is shown in Figure 2.2. A part of the TTG and GG rocks of the Murmansk complex, which are characterized by sharply differentiated REE patterns and devoid of appreciable Eu anomalies, is comparable to the typical Archean TTG (Martin, 1986, 1994; Martin et al., 2005). At the same time, certain differences from standard TTG associations are noted. Primarily, this is enrichment in potassium (K), especially in highly silicic rocks. Both TTG and GG rocks (Fig. 2.2) somewhat differ in (La/Yb)N ratios, although overlapping is also noted. In most TTG rocks, this ratio ranges from 40 to 70; melanocratic tonalite is characterized by (La/Yb)N below 40; the highest value, up to 215, is inherent to granodiorite. The (La/Yb) values in the GG series uniformly cover a range of 20–175. N It is known that REE patterns of typical TTG rocks are devoid of Eu anomalies, or this anomaly is insignificant (Martin, 1994; Smithies, 2000; Martin et al., 2005; Moyen and Martin, 2012). In contrast, an Eu anomaly (Eu/Eu* = 0.34–1.52) is an ordinary attribute of the rocks belonging to the Murmansk complex. The anomaly slightly increases with rising (La/Yb)N ratio. In the GG rocks, the highest Eu/Eu* (5.0) has been fixed in ultrasilicic highK granite. The rest of the data points fill a triangular field, the base of which approximately coincides with Eu/Eu* values in rocks of the TTG series. It is clearly seen that the arrangements of data points corresponding to the TTG and GG series in this plot

21

are quite different. The plots shown in Figure 2.3 demonstrate a sharp bend in REE redistribution in TTG and GG rocks. The fields of TTG and GG data points merge with an overlap near Eu/Eu* = 1. The sense of this feature, as well as orientation of the arrows that point to a suggested trend of the process, will be discussed later herein. Geodynamic Setting, Age, and Formation Conditions of the Granite-Migmatite Complex of the Murmansk Microcontinent The origination and evolution of the granite-migmatite complex can be reconstructed, to a significant degree, on the basis of geochemical data, especially REE patterns. The REE patterns with characteristic positive Eu anomaly in the early Precambrian TTG and GG complexes of the eastern Fennoscandian Shield, some of which were formed or metamorphosed under granulitefacies conditions, and in the rocks of the enderbite-charnockite series are well known (Lobach-Zhuchenko et al., 1984; Mints et al., 1996, and references therein; Glebovitsky, 2005; Mikkola et al., 2012); numerous examples are given in the subsequent chapters. Similar geochemical features in the early Precambrian TTG granitoids in other regions and in some younger TTG units also

Figure 2.2. Chondrite-normalized rare earth element (REE) patterns in rocks of the Murmansk granite-migmatite complex in the Dalnie Zelentsy area, modified after Mints et al. (1996). (1) amphibolite, (2) trondhjemite, (3) microcline-plagioclase granite; (4) plagioclasemicrocline granite. SiO2 contents are indicated in the figure. REEs were determined with instrumental neutron activation analysis (INNA). Compositions of C1 chondrite were taken from Taylor and McLennan (1985). See text for explanation.


22

Chapter 2

Figure 2.3. Positive Eu anomalies in rocks of the Murmansk granite-migmatite complex. Analytical data were taken from Mints et al. (1996). White dots—amphibolite; petrochemical granitoid series: gray dots and arrows—tonalite-trondhjemite-granodiorite (TTG); black dots and arrows—granite-granodiorite (GG). K2O, CaO, Al2O3, and SiO2 are in wt%; La and Yb, chondritenormalized concentrations, and chondrite composition are after Sun and McDonough (1989). See text for explanation.

should be mentioned (Knudsen and Andersen, 1999; Gasquet et al., 2003). At the same time, the currently popular models of TTG formation by partial melting of mafic rocks, likened to a certain degree to adakite-type volcanic rocks, do not mention this geochemical feature (Smithies et al., 2003; Martin et al., 2005). Meanwhile, specific REE patterns with moderate and low concentrations of both HREEs and LREEs, and with a pronounced Eu maximum against this background, were established for the initial melts making up a neosome of migmatite complexes. Such REE patterns close to those in feldspars are charac-

teristic for the initial portions of melts, which have solidified as a neosome of migmatites formed due to partial melting of rocks under conditions of upper-amphibolite or granulite facies. In certain cases, these melts are squeezed out, owing to filter pressing, to form large leucogranite bodies with the same geochemical features. Such a situation was studied in various regions, primarily for migmatized graywackes (Jung et al., 1998, 1999, 2000; Fornelli et al., 2002; Sheppard et al., 2003) and migmatized rocks of the TTG series (Guernina and Sawyer, 2003; Glebovitsky, 2005). Similar results were obtained in experiments on partial


Mesoarchean Kola-Karelia continent melting of mafic granulites (metabasalt and metagabbro) under deficiency in water (Springer and Seck, 1997). A positive Eu anomaly was fixed largely in Archean rocks; however, similar data on younger rocks are known as well. For example, positive Eu anomalies were noted in the products of partial melting of Hercynian granulites (Janoušek et al., 2000). An alternative interpretation of positive Eu anomalies based on fitting of REE patterns characteristic of partial melting–crystal fractionation assumes that trondhjemitic migmatites with unusually low REE contents and high positive Eu anomalies can be a quartz-plagioclase cumulate formed by crystallization of the initial trondhjemite melt (Bogatikov et al., 2006). A conclusion close in sense has been drawn for the migmatites formed as products of partial melting of the granulite Varpaisjärvi complex (section 3.1.4). It is suggested that the quartz-plagioclase assemblage with specific REE pattern and Eu maximum was formed at the early stage of felsic melt crystallization (Nehring et al., 2009). At the same time, it is evident that the model of quartz-plagioclase cumulate does not explain the origin of the frequently observed leucosome enriched in K-feldspar. In our opinion, a more viable genetic model assumes that felsic K-Na melts with positive Eu anomalies are a product of dry high-temperature (granulite- or high amphibolite-facies) partial melting of the previously formed mafic-to-felsic crustal rocks. A positive Eu anomaly in the eutectic is directly related to the predominance of plagioclase and K-feldspar in the melt. Large bodies of granitic melts with specific REE patterns are formed as a result of filter pressing. With heating of melt above the cotectic temperature, the Eu anomaly is leveled. Thus, the geological relationships and geochemical features of the Murmansk granitoids can be regarded to result from a subsequent melting event: TTG melted to produce GG rocks. Mikkola et al. (2012) arrived at a similar conclusion as a result of geochemical simulation of the formation of Neoarchean leucogranite from the Kianta complex. The available geochronological data (Table 2.1) allow us to reconstruct a sequence of events in the geological history of the Murmansk microcontinent. These data show that the ancient crust was intruded by juvenile magma both primitive (gabbroanorthosite) and evolved (enderbite, tonalite) from 2.93 to ca. 2.80 Ga. The high-temperature magmatic activity was accompanied by granulite-facies metamorphism of the crust. These events were followed by migmatization and granite formation as their immediate continuation. The Teriberka-type granitic rocks exposed in the central, probably more eroded part of the Murmansk microcontinent retained relict attributes of granulitic substrate. The Iokanga-type granitic rocks situated far from the central part of the Murmansk microcontinent were formed after amphibolite-facies rocks and are younger in age (2.77–2.72 Ga). The latest events of granite and pegmatite formation took place ca. 2.72–2.62 Ga. The fast transition from conditions of granulite facies to granite formation makes it possible to consider granulite metamorphism and generation of granitic magma as links in a single chain of processes. In general, a scenario characterizing forma-

23

tion of the Murmansk migmatite-granite complex can be represented as the following succession of events: (1) 2.93–2.77 Ga: underplating of an ancient crust with unknown properties by mafic magmas and intrusion (intraplating) of mafic magmas into this crust; creation of the conditions corresponding to granulite-facies metamorphism; partial melting of the mafic crust with formation of enderbite and trondhjemite melts; squeezing out and ascent of these hot melts into the upper crust; and magmatic replacement of country rocks with appearance of more mafic varieties of TTG series (geochemical characterization of the process is given in Fig. 2.3); and (2) from 2.83–2.80 to 2.77–2.72 Ga: new thermal events and related partial melting of the TTGs with appearance of potassic granites; squeezing out and transfer of K-enriched felsic melt; and magmatic replacement of country rocks with formation of GG association (geochemical characterization of the GG stage is given in Fig. 2.3). With allowance for structural features of the Murmansk microcontinent and by analogy with synforms of the Volgo-Uralia region (chapter 5), it is suggested that the tectonic depressions were formed under extensional conditions in the crust above local plumes. The arcuate and ring normal faults at the boundaries of these depressions drained sources of granodiorite-granite (GG) magmas and determined the localization of the corresponding intrusive bodies (Fig. 2.1). In contrast to the Volgo-Uralia region, the volcanic and sedimentary fill of the depressions has not been retained at the present-day denudation level. 2.1.2. Inari-Kola Microcontinent The Inari-Kola microcontinent (a part of the northern KolaKarelia continent) constitutes most of the Kola Peninsula. Its northeastern boundary coincides with the Kolmozero-Voronya suture zone (greenstone belt), which extends along the southwestern boundary of the Murmansk microcontinent. The southern boundary of the Kola microcontinent is marked by the Central Belomorian suture (greenstone belt). It is customary to define the Inari-Kola microcontinent as a granite-greenstone domain (Mints et al., 2010b, and references therein). The age and geodynamic setting of the basement complex, formed mainly by TTG rocks and greenstone belts, are different. Therefore, they are considered separately in the corresponding sections of this book. The Paleoproterozoic Pechenga-Imandra-Varzuga sedimentary-volcanic belt divides the Inari-Kola microcontinent into two domains: the northern Norwegian-Kola Province (block, microcontinent) and the southern province, consisting of the Inari and Tersky (Tersky-Strelna) blocks, which have been referred to as a part of the Kola Province by some authors and as a part of the Belomorian Province by others (see Glebovitsky, 2005, and references therein). For a long time, the Norwegian-Kola block was considered to be the oldest folded core of the Fennoscandian Shield. It was suggested that this core is largely composed of granulite-facies rocks and partly of granite-greenstone associations. Later on, geochronological


24

Chapter 2

and geophysical research, as well as drilling of the Kola Superdeep, compelled us to revise this concept. The microcontinent is composed of a more or less continuously traced rock association, including the area overlapped by thrust-nappe assembly of the Central Kola granulite-gneiss belt. At the present-day denudation level, the microcontinent consists of the fragments separated by nappes of the Central Kola belt and Paleoproterozoic sedimentary and volcanic structures (Appendices I-4 and I-5 [on CD-ROM accompanying this volume]). At the surface, the Inari-Kola microcontinent is composed of the granite-greenstone continental crust. The granitoid matrix is a combination of TTG gneisses and bodies of microcline-plagioclase granitoids formed synchronously with volcanic-sedimentary sequences of greenstone belts or soon after their origination. In this section, the TTG gneiss complexes are regarded as the basis of the Inari-Kola microcontinent. The basement complex is the best studied in the northeastern part of this microcontinent. According to Siedlecka et al. (1985) and Dobrzhinetskaya et al. (1995), the complex is composed of migmatized TTG gneisses separated by tectonic boundaries. The NW-trending gneissic banding and lineation dip to the northeast throughout the domain. The TTG gneisses of particular tectonic sheets, varying in composition and structure, have proper names: Varanger, Kirkenes, Svanwick, and a number of less important subdivisions. The characterization of TTG gneisses given herein was taken mainly from Dobrzhinetskaya et al. (1995). The basement complex underlying the Pechenga structure was penetrated by the Kola superdeep (Kozlovsky, 1984; Mints et al., 1996; Vetrin et al., 2002, and references therein) and is exposed in the eastern framework of the Pechenga structure. The Varanger complex is made up of alternating homogeneous and migmatized banded tonalite and trondhjemite gneisses. Some bands, which consist of paragneiss, amphibolite, and sporadic magnetite quartzite, occasionally extend for a few kilometers, having a limited thickness. Lenticular inclusions of serpentinized ultramafic rocks up to 1 km in extent are sparse. The rock composition is typically TTG series and characterized by sharply differentiated REE patterns, as a rule, without Eu anomalies; positive Eu anomalies are noted sporadically. The high-silicic trondhjemites reveal a negative Eu anomaly. The Kirkenes and Swanwick complexes are similar in composition. It is commonly suggested (without a special discussion) that TTG gneisses are mostly metamorphosed plutonic rocks. For the Archean TTG gneisses penetrated by the Kola superdeep, it was suggested that TTG gneisses are enderbite series metamorphosed under conditions of granulite facies (Duk et al., 1989); however, this suggestion was not confirmed by subsequent investigations (Chen et al., 1998). The study of inclusions in zircon opens new opportunities (Thomas et al., 2003). The primary melt inclusions have a glass phase in cores of prismatic zircon crystals from the Archean TTG in the section of the Kola superdeep (depth interval of 7622–12,262 m; Chupin et al., 2006, 2009), providing unequivocal evidence for the volcanic origin of the gneiss protolith. The

melt inclusions in central and intermediate parts of zircon cores from TTG at depths of 11,708–12,262 and 10,601–11,411 m correspond to rhyodacite-rhyolite (trondhjemite) composition, whereas the glasses located closer to core margins fit rhyolite and silexite. Findings of liquid CO2 syngenetic to melt inclusions show that early zircons crystallized in a deep magma chamber under dry conditions. The cores of the studied zircons crystallized from 2.88 to 2.83 Ga and 2.89 to 2.81 Ga at the two studied depth levels, correspondingly. Rhyodacite-rhyolite melts were entrapped during almost the entire time interval of core crystallization, whereas rhyolitic melts were captured mainly at the final stage of crystallization. The 207Pb/206Pb age of crystals with inclusions of dense CO2 (2.88 ± 0.02 Ga and 2.87 ± 0.01 Ga) is close to the age of early zircons with plagiorhyodacite-plagiorhyolite melt inclusions. The authors of this study separated the early deep-seated zircon crystals with plagiorhyodacite-plagiorhyolite melt inclusions and syngenetic inclusions of dense CO2 and the later zircons of volcanic origin. The Mesoarchean TTG gneiss complexes in other parts of the Inari-Kola microcontinent are fragmentary, in particular, because of widespread Neoarchean gneisses localized among the TTG gneisses of basement and related to younger events coeval with formation of greenstone belts (see chapter 3). Nevertheless, evidence for occurrence of TTG gneisses in basement is noted more or less constantly everywhere: along the boundary with the Murmansk microcontinent; in the eastern framework of the Paleoproterozoic Pechenga structure; in the Inari-Allarechka area; and near the eastern pinch-out of the Paleoproterozoic Imandra-Varzuga belt and its southern framework, where trondhjemite gneisses are predominant. The southwestern part of the microcontinent is composed of rocks pertaining to the Ena-Gridino granite-gneiss complex (hereafter we use term granite gneiss to denote a basement rock complex composed of TTG and GG, unspecified). This domain was conventionally interpreted as the tectonic Keret Nappe, consisting of paragneisses and granite gneisses, or as the northeastern margin of the Khetolambina terrane. In both versions, this granite-gneiss complex was involved into the Belomorian Province (Glebovitsky, 2005, and references therein). In the last decade, significant corrections have been introduced into this concept. In particular, it has been shown that the Keret Nappe and the Khetolambina microcontinent are separated by the Central Belomorian greenstone belt (suture), so that the Keret gneisses and granite gneisses can be regarded as the southern margin of the Inari-Kola microcontinent (Miller et al., 2005; Slabunov, 2008; Mints et al., 2010a, 2010b, 2014b). The southeastern part of the area occupied by the Ena-Gridino granite-gneisses, which are overlapped by the tectonic nappe of the Kolvitsa-Umba granulite belt and flooded by water of the White Sea, is fragmentary and exposed in the Karelian nearshore zone and on islands of the Kandalaksha Bay up to the settlement of Gridino (Appendices I-1 and I-4 [on CDROM accompanying this volume]). It was suggested for a long time that the granite-gneiss and the Central Kola granulite-gneiss complexes are connected by a



No. 1

2

3

4

TABLE 2.2. AGE DETERMINATIONS ON THE ARCHEAN GRANITOIDS IN THE INARI-KOLA MICROCONTINENT Complex, group Rock, process Age (Ga) Method References Varanger complex, Tonalite-trondhjemite2.87–2.80 U-Pb thermal Duk et al. (1989); including cores of granodiorite (TTG) gneiss ionization mass Balashov and Vetrin the Kola superdeep spectrometry (TIMS) (1991); Bibikova et al. zircon age (1993c); Chen et al. (1994, 1998); Levchenkov et al. (1995) Pegmatoid granite from core 2.93 U-Pb sensitive highBibikova et al. of the Kola superdeep resolution ion (1993c) microprobe (SHRIMP) zircon age U-Pb SHRIMP zircon Chupin et al. (2006, 2.80, Primary melt inclusions with age 2009) a glass phase in cores of 2.81, prismatic zircon crystals from 2.83 the Archean TTG in the section of the Kola superdeep, sequences from top to down: second, fourth, and tenth Same rocks, metamorphic 2.77–2.75 and rims around zircon crystals 2.70–2.67

5

Olenegorsk complex

TTG gneiss

6

Voche-Lambina complex

Trondhjemite gneiss

7 8

Granite-gneiss complex in the southeastern InariKola microcontinent

Trondhjemite pebble from conglomerate TTG gneiss

gradual transition. This idea was subject to revision when new geochronological data, geological mapping, and study of deep structure were undertaken. The Titovka granite-gneiss complex extends in the northern Inari-Kola microcontinent along the boundary with the Murmansk microcontinent (Appendix I-52). In the deep section along the 1-EU geotraverse (interval of 30–90 km), it is clearly seen that the Titovka complex is continuously traced toward the eastern framework of the Pechenga structure beneath the tectonic nappe of the Central Kola granulite-gneiss belt (see section 12.2.2; Appendix I-19 [on CD-ROM accompanying this volume]). At the same time, the results of geological mapping indicate that the granite-gneiss Titovka complex plunges in a northeastern direction beneath the Murmansk microcontinent. In general, the granite-gneiss Titovka complex is a fragment of the Inari-Kola microcontinent and in structural terms is a swell-like uplift, ~30 km in width, striking in the northwestern direction. At the surface, the Titovka and the East Pechenga complexes are separated by a synformal tectonic nappe of the Central Kola granulitegneiss complex.

2

25

Appendix I-5 appears on CD-ROM and at the end of this volume.

2.79

U-Pb TIMS zircon age

Bayanova et al. (1998) Mitrofanov and Pozhilenko (1991); Balashov et al. (1992) Kislitsyn et al. (2000)

U-Pb NORDSIM (Nordic ion microprobe facility) zircon age

Daly et al. (2001, 2006)

2.81–2.76

2.81 2.72–2.67

The ages of most TTG gneisses are estimated within a time interval of 2.87–2.79 Ga and to 2.76 Ga at the Voche-Lambina site in the central Inari-Kola microcontinent (Table 2.2). A single older estimate (2.93 Ga) has been obtained for pegmatoid granite from core of the Kola superdeep. The model Nd age, t(DM) (2.92– 2.82 Ga; εNd(t) = –0.6 to +2.68; Timmerman and Daly, 1995; Vetrin et al., 2002), is close to the time of magmatic crystallization and confirms the dominant juvenile character of magma and a low level of crustal contamination. According to Chupin et al. (2006, 2009), magmatic zircons with melt inclusions from tonalite-trondhjemite gneisses (metavolcanic rocks) in the Archean part of the Kola superdeep section have the following ages (from top down): 2.80 Ga for the second sequence; 2.80 Ga for the fourth sequence; 2.81 Ga for the eighth sequence; and 2.83 Ga for the tenth sequence. The glass in melt inclusions is evidence for an initial volcanic origin of the studied TTG gneisses (e.g., Thomas et al., 2003). Thus, the felsic volcanics in the studied part of the Archean section were formed during ca. 30 Ma. The U-Pb zircon age of the granite gneiss complex in the southeastern Inari-Kola microcontinent falls into interval of 2.72–2.67 Ga (Table 2.2). Much older Rb-Sr estimates for wholerock samples have been published. The Rb-Sr isochron yielded


26

Chapter 2

2.87 Ga; the Sr model age is 2.83 Ma (Balashov et al., 1992). The Nd model ages t(DM) are 3.04 and 2.95 Ga (Daly et al., 2001). These discrepancies probably indicate nonuniform assimilation of older crustal material by magmas, or, as was assumed by Balashov, it is caused by the residence time of TTG granitoids as early as 2.87 Ga in combination with metamorphism that took place ca. 2.72–2.67 Ga. The geochronological data, which characterize TTG gneisses localized along the southwestern boundary of the InariKola microcontinent (section 2.4), allow us to distinguish two rock groups: (1) TTG gneisses with predominance of tonalite dated at 2.82–2.77 Ga, and (2) granitoids of various composition with single estimates at 2.74, 2.69, and 2.63 Ga. Geodynamic Setting and Formation Conditions of the Granite-Migmatite Complex of the Inari-Kola Microcontinent The succession of main events in the history of the InariKola granite-greenstone domain can be characterized as follows (Table 2.2). The pregreenstone granite-gneiss complex, which is studied in full only in the northwestern part of the microcontinent, is composed largely of TTG series. It can be said with a high degree of confidence that the protolith of the TTG gneisses was mostly formed ca. 2.89–2.81 Ga as a product of volcanic eruptions of high-temperature rhyodacite-rhyolite magmas saturated with CO2. Crystallization of the volcanic rocks is dated at 2.83– 2.80 Ga. The volcanic or more probable volcanic-plutonic protolith of the pregreenstone tonalite-trondhjemite complex was created roughly as early as 2.80 Ga. With allowance for dry melting of mafic substrate, the geodynamic setting of magmatic activity hardly was related to subduction. An intracontinental plume-type setting, with intense heating, underplating by mafic magma, and subsequent melting of mafic rocks with formation of TTG melts, is more probable. As judged from so-far scant information, the time of trondhjemite gneiss formation in the eastern part of the microcontinent can be estimated at 2.87–2.83 Ga, and that in the southeastern part can be estimated at 2.82–2.77 Ga. Assuming a volcanic scenario for TTG gneiss origin, we have to explain the cause of crystallization of rather abundant and large zircon crystals in felsic volcanic rocks. It is suggested that volcanic eruptions were related to vigorous catastrophic ejections of foamed pyroclastic flows (flows of volcanic ash). As is known for recent pyroclastic flows, the degree of magma chamber crystallization before eruption could have been rather high, and pyroclastic flows (welded tuffs and ignimbrites) are compared with granitic plutons burst through to the surface.

width. Attention to the Vodlozero microcontinent (block, according to conventional nomenclature) grew sharply after the appearance of geochronological data on the older (>3.0 Ga) granitoids and greenstone rocks as compared with other parts of the craton (Sergeev et al., 1990a; Lobach-Zhuchenko et al., 1993b). The western boundary of the Vodlozero microcontinent is marked by the Vedlozero-Segozero system of greenstone belts. The Sumozero-Kenozero system plays a similar role in the east (Appendix I-4). Information on the northern boundary of the Vodlozero microcontinent is less definite. The accepted concept (Appendices I-1 and I-4 [on CD-ROM accompanying this volume]) is based on the results of geological mapping and interpretation of seismic images of the crust along the 1-EU geotraverse. A similar view on this boundary has been accepted by Kozhevnikov (2000) and Svetov (2005). In Glebovitsky (2005), the boundary is shown somewhat southerly, i.e., immediately to the north of the Shilos greenstone belt.

2.1.3. Vodlozero Microcontinent

The Older Pregreenstone Granite-Gneiss Complex Tonalite-trondhjemite gneisses are found in several areas. Medium- and coarse-grained tonalite-trondhjemite (TT) gneisses are predominant. Stratiform bodies of fine-grained gneiss, the same in composition and a few tens of centimeters to a few meters in thickness, are second in abundance. They form up to 10% of outcrops among the coarse-grained TT gneisses, with sporadic inclusions of amphibolites and meta-ultramafic rocks. Some amphibolites close in composition to high-Mg and high-Si basalts are interpreted as products of assimilation of felsic crustal rocks by komatiitic melt (Arndt et al., 1997). The coarse- and fine-grained TT gneisses correspond to highAl tonalite and trondhjemite with low Rb, Ba, and REE contents; fractionated REE patterns with positive Eu anomalies; sharp U and Nb minimums; and positive Sr and Zr anomalies (Bogatikov et al., 2006). The positive Eu anomaly against the background of strongly fractionated REE patterns, especially in highly silicic varieties, can indicate creation of the TT protolith as a result of high-temperature, granulite- or high amphibolite-facies melting of the previously formed TTG-type crust (see section 2.1.1 for more details). The ancient crust should have significant thickness to be melted in equilibrium with garnet- and amphibole-bearing restite. The upper temporal boundary of the older (pregreenstone) gneisses is determined by the Lairuchei pyroxenite-gabbro-diorite intrusion (Chekulaev et al., 1994) and diorite dikes. Variable MgO and Al2O3 contents in ultramafic and mafic rocks (8.3–18.2 wt% and 20.0–13.7 wt%, respectively) and a decrease in initial values of εNd from –0.8 to –1.1 to –2.6 to –2.7, with an increase in silica content, suggest that the variation in composition could have been related to contamination of mantle-derived magma with older crustal matter (Kulikov et al., 1990; Bogatikov et al., 2006).

The Vodlozero microcontinent occupies most of the area in the southeastern half of the Karelian craton. Taking into account its continuation beneath sedimentary cover of the East European Platform, the total extent reaches 550–600 km and ~400 km in

Pregreenstone Biotite and Biotite-Amphibole Plagiogneisses and Amphibolites Pregreenstone biotite and biotite-amphibole plagiogneisses and amphibolites are rather abundant in the Vodlozero


Mesoarchean Kola-Karelia continent microcontinent along with granite gneisses (Lobach-Zhuchenko et al., 1989). As a rule, these rocks had undergone intense tectonization and migmatization, and because of this, the relationships between older and younger rocks became significantly complicated. In the opinion of Lobach-Zhuchenko et al. (1989), the gneiss complex can be regarded as an association of metavolcanic rocks, the composition of which varies from basaltic andesite to dacite. The geochemical characteristics provide evidence for a moderate depth of magma generation, and positive εNd initial values of +0.7 to +2.9 indicate the juvenile nature of magma with an insignificant level of crustal contamination. At the same time, initial εNd of amphibolites ranges from +3.0 to –1.2 and thus points to assimilation of the older crustal material by mafic magma. The first evidence for occurrence of TTG and biotiteamphibole gneisses older than 3.0 Ga in the Vodlozero microcontinent (block; Sergeev et al., 1990a; Lobach-Zhuchenko et al., 1993b) has been recently supported and supplemented. Using conventional and sensitive high-resolution ion microprobe (SHRIMP) II technology for zircon dating, it has been shown that the igneous protolith of the TTG gneisses was formed 3.24– 3.13 Ga (Table 2.3). The cores of zircon crystals often contain fluid and completely crystallized melt inclusions, as well as inclusions of orthoclase, orthopyroxene, quartz, ilmenite, and galena. The inclusion study shows that autometasomatic processes developed at a high CO2 activity in fluid (Sergeev et al., 2008). The U-Pb SHRIMP II age of zircon from biotite-amphibole gneiss estimated at 3.24–3.15 Ga completely overlaps the age of TTG gneisses. The date of 3.24 Ga is interpreted as a protolith age of the amphibolite-gneiss complex, which underwent first metamorphism ca. 3.15 Ga contemporaneously with emplacement of tonalite plutons. The fine- and coarse-grained TTG gneisses have model Nd age of 3.4 Ga; the value of εNd (= 1.5–2.6) recalculated to 3.2 Ga indicates juvenile mantle or a mafic lower-crustal source of melts. Ultramafic inclusions, which are suggested to be metakomatiites, are also dated at 3.4 Ga (Sm-Nd whole-rock isochron), εNd = 1.2 (Puchtel et al., 1991). The model Nd age of biotite-amphibole gneiss is 3.3–3.2 Ga and coincides with the zircon age. The model Nd age of amphibolite is 3.5–3.2 Ga (Lobach-Zhuchenko et al., 1993b). The U-Pb zircon age of the Lairuchei pyroxenite-gabbro-

No. 1 2 3 4

5

27

diorite intrusion in the northeastern Vodlozero microcontinent determines the upper temporal boundary of tonalite-trondhjemite gneisses as 2.976 Ga. The same age (2.97 Ga) has been obtained for amphibolite dike of the earlier generation (Table 2.3). Geodynamic Setting and Formation Conditions of the Granite-Gneiss Complex of the Vodlozero Microcontinent The main rock types of the Vodlozero microcontinent have comprehensive and detailed, though not always mutually coherent, characteristics. The consideration of available information allows us to interpret the geodynamic settings and the succession of main events in the geological history of the microcontinent. The pregreenstone association of TTG gneisses, paragneisses, and amphibolites was formed 3.24 to ca. 3.0 Ga. The data, which make it possible to discuss the origin of the ancient crust, are limited and partly controversial. The initial εNd estimates, as a rule, indicate the juvenile character of the rocks, except for protoliths of amphibolites, the parental melts of which apparently have a mixed mantle-crustal origin. Possible involvement of the ancient crust in magma formation is supported by an old model Nd age. The negative εNd values, which decrease from –0.8 to –1.1 to –2.6 to –2.7 with growth of silica content, also show a mantle-crustal nature of the Lairuchei intrusion and involvement of much older continental crust in magma generation. The REE patterns with frequent positive Eu anomalies in TTG gneisses and occurrence of CO2 fluid inclusions in cores of ancient zircon crystals can be regarded as evidence for tonalitetrondhjemite gneiss formation as a product of high-temperature (granulite- or high amphibolite-facies) melting of the older, thick TTG crust. 2.1.4. Ranua and Iisalmi Microcontinents According to geochronological data, the Ranua and Iisalmi microcontinents in the western and northwestern parts of the Karelian craton (Appendices I-1 and I-4) are fragments of the Paleoarchean and Mesoarchean TTG crust. Their localization in the western Karelian craton assumes certain commonality in geological evolution of these crustal blocks (Mints et al., 2010b); however, the U-Pb, Lu-Hf, and Sm-Nd isotopic geochronological

TABLE 2.3. AGE DETERMINATIONS ON THE ARCHEAN IGNEOUS ROCKS IN THE VODLOZERO MICROCONTINENT Complex, group Rock, process Age (Ga) Method References From 3.24 U-Pb sensitive highSergeev et al. (2007a, Pregreenstone Igneous protolith of tonaliteto 3.13 resolution ion 2008); Chekulaev et granitoid complexes trondhjemite-granodiorite microprobe al. (2009) (TTG) gneisses (SHRIMP) zircon age 3.24 Sergeev et al. Igneous protolith of the (2007a) amphibolite-gneiss First metamorphism of the 3.15 amphibolite-gneiss Lairuchei intrusion Pyroxenite-gabbro-diorite 2.76 U-Pb thermal Chekulaev et al. ionization mass (1994) spectrometry (TIMS) zircon age M afic dike Am phibolite 2.97 U - P b S H R IM P z i r c o n Chekulaev and age Arestova (2009)


28

Chapter 2

data show that before ca. 2.8 Ga, the Karelian complexes could have evolved separately (Lauri et al., 2011). In the east, the Ranua and Iisalmi microcontinents border on the Kianta microcontinent. Having been formed in the Paleoproterozoic, the transpressional boundary is a system of right-lateral strike-slip and thrust faults, along which the Ranua and Iisalmi microcontinents were offset to the north and thrust eastward over the Kianta microcontinent. The boundary fault zone in its central segment is a packet of tectonic sheets composed of rocks pertaining to the Paleoproterozoic Kainuu sedimentary-volcanic belt, which is thrust over granite-greenstone crust of the Kianta microcontinent, in line with general structure (see section 12.3.4). In the west and southwest, the Ranua and Iisalmi microcontinents border on the Paleoproterozoic Svecofennian accretionary orogen. The northern margin of the Ranua microcontinent is overlapped by a Paleoproterozoic volcanic-sedimentary sequence that fills the Peräpohja structure and is thrust to the south. The microcontinent contours coincide with contours of the Iisalmi and Ranua terranes in terms of Sorjonen-Ward and Luukkonen (2005) separated by the NE-trending Oulujärvi strike-slip fault zone (Appendices I-1 and I-4). Ranua Microcontinent The Ranua microcontinent is a block of the Archean crust, triangular in plan view, in the northwestern Karelia craton. Because of insufficient outcrops and monotonous trondhjemitetonalite gneisses, the geology of this area remains the worst studied in Finland. The Ranua microcontinent involves the Kalpio paragneiss belt and the Oijärvi greenstone belt, as well as the Pudasjärvi granulite complex, which was called the “Siurua complex” by Sorjonen-Ward and Luukkonen (2005). We will use the term “Siurua” to denote the TTG complex and retain the name “Pudasjärvi” for the granulite complex following the earlier publication by Mutanen and Huhma (2003). The age of magmatic crystallization of the trondhjemitegneiss Siurua complex is estimated at ca. 3.5 Ga (U-Pb NORDSIM (the Nordic ion microprobe facility) zircon age). To date, the Siurua complex is the oldest known in the Fennoscandian Shield. The model Nd age is practically the same (3.48 Ga). The εNd (2700) is a very low (–10.8) and corroborates the ancient age of the gneiss. Zircons point out several events in the history of the Siurua complex, mainly within a range of 3.5–3.4 Ga. One of the first ages, the age of the core in a zircon crystal, was estimated at 3.73 Ga (Mutanen and Huhma, 2003). Later on, sparse xenogenic cores of zircon crystals from tonalitic and leucogranitic rocks yielded ages of 3.6–3.7 Ga, and initial 176Hf/177Hf suggested the influence of a crustal component with t > 4.0 Ga (assuming a chondritic uniform reservoir [CHUR]–like mantle source; Lauri et al., 2011). Magmatic zircons in a felsic granulite from the Pudasjärvi granulite belt have a U-Pb age of ca. 2.96 Ga (Mutanen and Huhma, 2003), indicating a time of epicontinental felsic magmatism activity. The ca. 2.7 Ga high-grade metamorphism within Pudasjärvi belt, which is related to a Neoarchean stage of crustal evolution, will be discussed in chapter 3 (section 3.1.4).

Iisalmi Microcontinent The Iisalmi microcontinent is the southwestern part of the Karelian craton (the Iisalmi terrane after Sorjonen-Ward and Luukkonen, 2005). A significant part of the Iisalmi crust consists of tonalite-trondhjemite migmatized gneisses with enclaves of amphibolite and ultramafic rocks (Paavola, 1986, 1999). The eastern part of the Iisalmi microcontinent is regarded by Hölttä and coauthors (2012a) as the Rautavaara complex, primarily composed of TTG gneiss containing amphibolite and biotiteplagioclase paragneiss inclusions and schlieren. The abundance of chemically altered ultramafic to felsic rocks (cordierite orthoamphibolite and quartzite with Al-silicates and cordierite) is a distinctive feature of the Rautavaara complex (Hölttä et al., 2012a). Previously, Mänttäri and Hölttä (2002) regarded the Iisalmi and the Rautavaara areas as a single whole, assuming that the Rautavaara crust is composed of retrogressed rocks of the Iisalmi complex. The predominant part of the Iisalmi microcontinent is occupied by a synform of the Neoarchean Varpaisjärvi granulite-gneiss complex, which is characterized in chapter 3 (section 3.1.4). A quartz diorite paleosome of the migmatized TT gneisses yielded U-Pb zircon ages of 3.14 and 3.10 Ga (Paavola, 1986), making it one of the oldest rocks in the Karelia craton. The paleosome of migmatized granulite orthogneiss dated at 3.2 Ga (Hölttä et al., 2000; Mänttäri and Hölttä, 2002) is also among the oldest rocks in the shield. The signs of Mesoarchean metamorphic events are difficult to distinguish in the small enclaves of Mesoarchean rocks. Mänttäri and Hölttä (2002) interpreted a zircon population dated back to 3.1 Ga as evidence for metamorphism of the Iisalmi complex at ca. 3.2 Ga. According to refraction (DSS) and reflection (CMP) seismic profiling, the Iisalmi block is distinguished by unusually thick crust (55–60 km; Korsman et al., 1999; Kukkonen and Lahtinen, 2006; Kozlovskaya et al., 2004), in contrast to the ~40 km crust in the adjacent granite-greenstone domains of the Karelian craton. A seismic image of the crust along the FIRE-1 profile (see section 12.3 and Figures 12.6 and 12.7 for more details) shows that the tectonic sheets of the accretionary complex are traced from the surface near the town of Kiuruvesi (stake 275 km) to the town of Kuhmo (close to stake 75 km), where they reach the crust-mantle interface and, plunging further, “dissolve” in the mantle. The present-day Iisalmi crust above the accretionary complex is ~20 km thick. The Iisalmi crust is characterized by deep crustal xenoliths carried up by late Neoproterozoic kimberlite pipes in the Kaavi-Kuopio area (Appendix I-1 [on CD-ROM accompanying this volume]). Deep xenoliths are mainly composed of Archean and Paleoproterozoic mafic granulites. The U-Pb single-grain zircon age of ca. 3.5 Ga and the model Sm-Nd age t(DM) (ca. 3.7 Ga) of the xenoliths correspond to the oldest upper-crustal rock association. It was suggested that these rocks represent a major mantle-plume event responsible for formation of a significant body of Paleoarchean continental crust (Peltonen et al., 2006).


Mesoarchean Kola-Karelia continent 2.1.5. Kianta Microcontinent The Kianta microcontinent (in our understanding) is situated between the ancient Ranua and Iisalmi microcontinents in the west and the central Karelian craton, occupied by the Kuhmo-Segozero microcontinent, in the east (Appendix I-4 [on CD-ROM accompanying this volume]). The western boundary of the Kianta microcontinent is obscured by nearly meridional strike-slip fault zones, volcanic-sedimentary complexes of the Neoarchean Kalpio belt, and the Paleoproterozoic Kainuu belt (Sorjonen-Ward and Luukkonen, 2005; Kontinen et al., 2014). Its eastern boundary is marked by the fragmented TipasjärviKuhmo-Suomussalmi greenstone belt, which we interpret as an analog of the modern suture zones (structure and provenance of the Tipasjärvi-Kuhmo-Suomussalmi greenstone belt are considered in section 2.2.4). Following Sorjonen-Ward and Luukkonen (2005), who have noted that general similarities in structural architecture, age, and petrography of granitoids show that at least by 2.74 Ga, the Ilomantsi and Kianta terranes were a single coherent terrane, we consider the Mesoarchean history of the Ilomantsi terrane after Sorjonen-Ward and Luukkonen (2005), or Ilomantsi complex after Hölttä et al. (2012a, 2012b), to be a part of crustal evolution of the Kianta microcontinent as a whole. In contrast to our interpretation, Sorjonen-Ward and Luukkonen (2005) called a much more extensive area the “Kianta terrane,” and its eastern boundary was regarded as an inferred border between the Western Karelia Subprovince and Central Karelia Subprovince. Hölttä et al. (2012a, 2012b) called the same area the “Lentua complex.” The Tipasjärvi-Kuhmo-Suomussalmi greenstone belt, in the interpretation of these authors, is located in the central part of the Kianta terrane. Here, we do not go beyond a brief characterization of the TTG gneiss complex, which occupies almost the entire Kianta microcontinent at the present-day denudation level. The deep structure of this domain is discussed on the basis of the FIRE-1 profile (Kukkonen and Lahtinen, 2006) and is reproduced in our interpretation of the seismic image of the crust (see sections 12.3.3 and 12.3.4). The TTG complex is made up of migmatized rocks with numerous inclusions of banded amphibolites (Fe-rich metatholeiites), unspecified in age, which probably are relics of the older mafic crust. According to Mikkola et al. (2011a), the TTG association consists of the rocks with fractionated REE patterns sharply depleted in HREE and with a negative or positive Eu anomaly (Eu/Eu* = 0.68–1.65). The crustal TTG rocks are distinguished by low Mg numbers (38, on average) and usually by Cr and Ni contents, which are below the detection limits (30 and 20 ppm, respectively). This suggests absence of or limited interaction of the melts with mantle. The TTG magmatism can be divided into three major phases: ca. 2.95 Ga, 2.83–2.78 Ga, and 2.76–2.72 Ga. Geochemical features, inherited zircons (3.24–3.07 Ga), and the negative initial εNd values of most rocks indicate involvement of older crust in their formation. Pb/Pb isotopic data (Vaasjoki et al., 1999) and

29

Sm-Nd model ages (Käpyaho et al., 2006; Luukkonen, 2001) indicate participation of an older (>3 Ga) crust. The Paleoarchean (3.24–3.07 Ga) zircons and commensurable model Nd ages support the existence of Paleoarchean protocrust. This protocrust could have represented a previously unrecognized basement underlying the oldest volcanic rocks of the Tipasjärvi-KuhmoSuomussalmi greenstone belt. Nevertheless, it remains unclear whether any of the oldest rocks have survived to the present time or if they have all been recycled during subsequent events (Mikkola et al., 2011a). The oldest known TTG associations are ca. 2.95 Ga in age. Based on the Sm-Nd isotopic parameters, recycling of the earlier protocrust was involved in the generation of these rocks. There is a break of nearly 100 m.y. between the 2.95 Ga event and the next phase of crustal growth. The time interval 2.83– 2.78 Ga clearly represents the most voluminous TTG association in the northern Kianta microcontinent. In the Kuhmo area, Käpyaho et al. (2006) divided this interval into two episodes (2.83–2.81 Ga and ca. 2.79 Ga). When considering the northern part of Kianta as a whole, these two episodes seem to merge into a single protracted magmatic event between 2.83 and 2.78 Ga. The Mesoarchean continental crust of the Kianta microcontinent as a whole demonstrates a progressive evolution toward more granitic compositions as a result of subsequent melting of metabasalts to produce a TTG association, which, in turn, was subject to melting with formation of transitional TTG association and granites (Mikkola et al., 2012; see discussion in section 2.1.1 for comparison). 2.1.6. Kuhmo-Segozero Microcontinent The Kuhmo-Segozero microcontinent, which occupies the central part of the Karelian craton, has oval outlines in plan view. Its extent from the north southward is 400 km, and the width attains 180 km. The southern part of the microcontinent is cut off by the boundary of the Svecofennian accretionary orogen (Appendices I-4 and I-6 [on CD-ROM accompanying this volume]). The Kuhmo-Segozero microcontinent is regarded here as a separate tectonic unit for the first time. In lithology, this microcontinent is a typical granite-greenstone domain. As follows from geochronological and geological data, most greenstone sequences are younger than the granite-gneiss basement. The structural unity of this domain, localized between the Belomorian Province and the Vodlozero microcontinent in the east and between the Kianta and Iisalmi microcontinents in the west, is emphasized by the oval concentric arrangement of greenstone belts shifted to the microcontinent margins (Appendices I-4 and I-5). We attempted to provide insights into the internal structure of the Kuhmo-Segozero microcontinent on the basis of available geological maps and the results of three-dimensional (3D) simulation (see sections 12.1.1 and 12.3.4). The seismic image of the crust allows us to trace greenstone belts and their fragments to depth. In summary, the structure of the Archean crust in the central part of microcontinent in plan view is a gentle semi-oval


30

Chapter 2

synform. The crustal structure is a set of bowl-shaped sheets enclosed into one another and composed of granite gneisses intercalated by volcanic-sedimentary greenstone complexes and/ or their roots. In the southern Kuhmo-Segozero microcontinent, the relationships of the granite-gneiss complex with lacy centroclinal structural elements of paragneiss belts make it possible to outline another semi-oval depression. Inasmuch as the oval-concentric structure of the KuhmoSegozero microcontinent is distinctly crossed by Paleoproterozoic dikes and sedimentary-volcanic belts, it is evident that this structure is a result of Archean (pre-Paleoproterozoic) tectonics. The predominant portion (95% according to some estimates) of the present-day surface of the microcontinent is occupied by TTG gneisses. Relatively full information on the composition and age of these rocks is available only close to greenstone belts, whereas at a distance from them, geochemical and geochronological data on TTG gneisses remain fragmentary. The older TTG rocks in the northern part of the microcontinent are homogeneous, medium- to coarse-grained, high-Al trondhjemite gneisses with fractionated REE patterns sharply depleted in HREEs and with positive Eu anomalies (Eu/Eu* = 1.5–1.7; Samsonov et al., 2001). The zonal zircon crystals with older cores indicate probable participation of ancient crustal matter in petrogenesis. The positive Eu anomaly, which was left unnoticed by Samsonov et al., shows that trondhjemites were formed as products of high-temperature melting of mafic material (see comprehensive discussion in section 2.1.1). The oldest age of this granite-gneiss complex was obtained for a mesosome of migmatite 20 km to the east of the Kuhmo belt. The mesosome contains zircons with cores dated at 2.94 Ga. The mesosome age (age of metamorphism) is ca. 2.84 Ga. The outer rims of the same zircon crystals, which are dated at 2.81 Ga, probably are products of late metamorphism and migmatization (Käpyaho et al., 2007). The zircon age from the mesosome of other samples taken to the east of the Kuhmo and Suomussalmi belts is 2.84–2.80 Ga. A younger zircon age (2.79 Ga) was obtained ~30 km to the east of the Suomussalmi (Bibikova et al., 2005b). Samsonov et al. (2005) suggested that the positive initial εNd values (2.4–1.8) indicate that melts are products of a mafic source, probably with insignificant contribution of a subduction-related component. As follows from the aforesaid, the early crust in the Kuhmo-Segozero microcontinent is dated at 2.94–2.80 Ga. The transformation of this crust consisted of two groups of events. The first group includes the events that completed formation of the microcontinent and immediately followed one another ca, 2.81–2.79 Ga. The second group of events, which involved the vast continental domain of the amalgamated Karelian and Belomorian accretionary orogens, started ca. 2.76 Ga and finished ca. 2.70 Ga. In the regional aspect, the events of the second group correspond to one of the most important stages in the history of the Archean Kola-Karelia continent as a whole. Synthesis of the available data in the broad regional aspect is given in section 3.1.

2.1.7. Khetolambina Microcontinent For a long time, most researchers considered the Khetolambina granite-greenstone domain to be a sheetlike body limited in thickness, which can be regarded as a stratigraphic subdivision (Stenar, 1972) or a tectonic sheet (Miller and Mil’kevich, 1995). Recently, a model was proposed, according to which the Belomorian mobile belt was a zone of geodynamic interaction between the Karelian craton and Khetolambina terrane (Miller et al., 2005). The seismic image of the crust along the 1-EU geotraverse, which crosses the Khetolambina microcontinent within an interval of 450–750 km, makes it possible to trace structures of this microcontinent from the surface to the base of the crust (Appendix I-19 [on CD-ROM accompanying this volume]; see section 12.3.4). The Khetolambina microcontinent extends for 600 km within the Fennoscandian Shield and continues presumably for 500 km beneath the sedimentary cover of the East European craton, having a width of 50–100 km (Appendices I-1 and I-4 [on CD-ROM accompanying this volume]). The microcontinent consists of TTG-type granite gneisses, younger granitic rocks, and greenstone volcanic-sedimentary associations. Near the northwestern end of the microcontinent, the granite-greenstone complex is exposed at the crest of a granite-gneiss dome among rocks pertaining to the base of the Paleoproterozoic sequences pierced by this dome. The Paleoproterozoic lopolith-like Koitelainen and Keivitsa layered mafic-ultramafic intrusions have been pushed up to the upper crust by this dome. Thereby, the nature prepared the Tojottamanselkä window for the Archean granite-greenstone complex, which emerges through both the Koitelainen intrusion and the base of the Paleoproterozoic sequence (Kröner et al., 1981; Mutanen and Huhma, 2001; Appendix I-1). The Kiviaapa window, which is larger in size and situated nearby and almost in the center of the Koitelainen intrusion, makes it possible to see a small portion of Archean paragneisses (metatuffites and metasedimentary rocks; Fig. 2.4; Mutanen and Huhma, 2001). The Finnish geologists compare these Archean rocks to the Pomokaira terrane (Sorjonen-Ward and Luukkonen, 2005). The granite-greenstone complexes accessible for observation owing to these windows are conjugated in space and structure with the northwestern end of the Khetolambina microcontinent. This was a reason to combine the aforementioned complexes with the Khetolambina microcontinent. The granite-gneiss Khetolambina complex, composed of TTG, biotite gneisses, and biotite-amphibole gneisses, occupies most of the microcontinent at the present-day denudation level. Paradoxically, the granite gneisses remain poorly studied and only at sporadic local sites and without any system at all. These rocks were investigated more thoroughly in the northern part of this microcontinent near Lake Kovdozero and especially near the Tupaya Guba (Gulf) of Kovdozero (Slabunov, 2008; Glebovitsky, 2005, and references therein). The granite-gneiss complex mostly consists of high-Al TTG gneisses. In the southeastern Khetolambina microcontinent, granitoids are concisely characterized along


Mesoarchean Kola-Karelia continent the southern shore of the Onega Gulf. Biotite, biotite-amphibole, epidote-biotite, and amphibole gneisses, often garnet bearing, are typical of this district. Small tonalite and granodiorite bodies are hosted in these gneisses. To the point, the main body of substantial information on the granite-gneiss complex of the Khetolambina microcontinent has been published by geochronologists. Up to five generations of granite gneisses and granitoids, the most of which are referred to as TTG series, are distinguished on the basis of geochronological data. As a rule, this conclusion has been drawn from incomplete geochemical information and most frequently from petrographic attributes. Necessary geochemical criteria were used only in rare cases (Glebovitsky, 2005; Bogatikov et al., 2006; Slabunov, 2008, and references therein). Granitic rocks of the oldest generation were identified at the northeastern end of the Khetolambina microcontinent near the Koitelainen mafic-ultramafic layered intrusion. This intrusion and early Paleoproterozoic country rocks nearby make up a peculiar cap at the crest of the granite-gneiss dome, which apparently created this unusual brachyanticline (Mutanen and Huhma, 2001). In the Tojottamanselkä and Kiviaapa windows, 2 and 8 km across, respectively, which are situated in the central and western parts of the intrusion, the Archean tonalite gneiss crops out. These probably are middle-crustal rheomorphic granitoids, which protruded through the early Proterozoic volcanic

31

and sedimentary rocks together with mafic-ultramafic lopolith and are now exposed within its contour (Fig. 2.4). Tonalite gneiss in the Tojottamanselkä window are known, owing to the detailed geochemical studies carried out by Kröner et al. (1981), Jahn et al. (1984), and Kröner and Compston (1990). These authors were first to prove the rocks older than 3 Ga in the Fennoscandian Shield. Although several generations of granitic rocks in the Khetolambina microcontinent were known for a long time, the massifs that correspond to particular generations were not outlined, and their structural and geochemical features remained unknown. The U-Pb zircon age of tonalite exposed in the Tojottamanselkä window is 3.11 Ga (Table 2.4; Kröner et al., 1981). Later on, this estimate was corroborated by SHRIMP technology (3.12 Ga; Kröner and Compston, 1990). The older cores of zircon grains dated at 3.25–3.16 Ga were established using SHRIMP. The REE and Nd isotope data (Jahn et al., 1984) imply that the tonalite precursor was basaltic in composition. The age of migmatization is 2.85–2.82 Ga, and the age of metamorphism was estimated only approximately at 2.73 Ga. The age of the oldest granitoids in the framework of the Tulppio greenstone belt was obtained as a result of special geochronological studies carried out by the Geological Survey of Finland. The U-Pb zircon age of tonalite and granite from the gneiss dome pertaining to the Ahmatunturi is 2.83 Ga and 2.90 Ga, respectively (Table 2.4; Juopperi and Vaasjoki, 2001).

Figure 2.4. Geological map of the Koitelainen layered intrusion pierced by the Tojottamanselkä and Kiviaapa domes exposed in tectonic windows of the oldest Khetolambina granite-greenstone complex, after Kröner et al. (1981) and Mutanen and Huhma (2001). (1, 2) Archean granite-greenstone complex, ca. 3.1 Ga: (1) mainly tonalite gneiss and paragneiss, (2) granite; (3–6) early Paleoproterozoic: (3) arkoses, volcanic breccia, high-Mg basalt, basaltic komatiite, and high-Mg andesite, 2.53 Ga; (4–6) Koitelainen layered intrusion and a fragment of the late Paleoproterozoic Keivitsa layered intrusion near the southern margin of map: (4) ultramafic rocks, (5) gabbroids, (6) granophyre capping; (7) fault.


32

Chapter 2

No. 1

2 3 4 5 6 7 8 9

TABLE 2.4. AGE DETERMINATIONS ON THE PALEOARCHEAN–MESOARCHEAN GRANITOIDS IN THE KHETOLAMBINA MICROCONTINENT Complex, group Rock, process Age (Ga) Method References Tonalite, magmatic zircons 3.11 U-Pb thermal Kröner et al. (1981) ionization mass spectrometry (TIMS) zircon age 3.12 U-Pb sensitive highKröner and Compston resolution ion (1990) Tonalite, zircon cores 3.16 and Tojottamanselkä dome microprobe 3.25 (SHRIMP) zircon age Tonalite, migmatitic zircons 2.85–2.82 Tonalite 3.1 Sm-Nd model age Jahn et al. (1984) 2.9 Rb-Sr model age Kröner et al. (1981) Tonalite, m etam orphic 2.7 Rb-Sr isochron age reworking Ahm atunturi com plex

Tonalite Granite

Geodynamic Setting and Formation Conditions of the Granite-Gneiss Complex of the Khetolambina Microcontinent Attributes of pregreenstone granite-gneiss basement at the surface are known only in the extreme northwestern end of the Khetolambina microcontinent. This is the tonalite gneiss, 3.12–3.11 Ga in age, exposed in the Tojottamanselkä dome and characterized earlier herein. The zircon cores dated at 3.25– 3.16 Ga suggest the existence of two separate predated events in the Mesoarchean crustal evolution. The rocks of the continental crust formed 200–500 m.y. before melting had a magma source.

2.83 2.90

U-Pb TIM S zircon age

Juopperi and Vaasjoki (2001)

There is indirect evidence that some blocks of the ancient continental crust probably also existed in other parts of the microcontinent. Important information is deduced from the zircon grains incorporated into the Chupa granulite gneiss. The origin and sources of various zircon populations will be discussed in section 3.1.4. Here, we only note that ages of detrital zircons in the Chupa gneisses mark two early time intervals: 3.22–3.10 Ga and 3.06– 2.90 Ga (Bibikova et al., 2004). These data furnish evidence in favor of the existence of ancient continental crust in the framework of the sedimentary basin or immediately in the basement of the Khetolambina microcontinent. It is impossible to estimate the location and dimensions of these crustal blocks from the available data.



33

2.2. MESOARCHEAN GREENSTONE BELTS: ISLAND-ARC SYSTEMS, ACCRETIONARY OROGENS, AND MATURE ISLAND ARCS (MICROCONTINENTS) Michael V. Mints

Geological, geochemical, and geochronological data allow us to recognize several greenstone belts, which are fragments of paleo-island-arc systems accreted to the margins of ancient microcontinents (continental nuclei): the Kolmozero-Voronya, Central Belomorian, Vedlozero-Segozero, Sumozero-Kenozero, and Tipasjärvi-Kuhmo-Suomussalmi belts; and the mature island arcs (microcontinents): Khetolambina and Kovdozero. These structural units are characterized by significant extent, close to rectilinear trend, localization along the boundaries between Archean microcontinents, and a specific set of petrotectonic assemblages. 2.2.1. Vedlozero-Segozero System of Greenstone Belts The greenstone belts in the framework of the Vodlozero microcontinent are divided into two groups. The VedlozeroSegozero system of NNE-trending greenstone belts extends along the western boundary. The northern end of this system, i.e., the South Vygozero belt, strikes in the latitudinal direction. As is shown by a model of deep structure along the 1-EU geotraverse (see Fig. 12.7), the sheetlike bodies composed of greenstone and granite-gneiss complexes alternate in the VedlozeroSegozero system of belts. They plunge eastward beneath the crust of the Vodlozero microcontinent. The second, SumozeroKenozero system of greenstone belts follows along the northeastern boundary of the Vodlozero microcontinent, cutting off the South Vygozero belt. The dip in the northeastern direction beneath the Khetolambina microcontinent is characteristic of this system. The mafic (amphibolite) dikes within the Vodlozero microcontinent are comparable in composition and age with particular parts of greenstone belts (Glebovitsky, 2005; Chekulaev and Arestova, 2009). The Vedlozero-Segozero system of greenstone belts extends for ~200 km, having a width of 50–60 km (Appendix I-4 [on CD-ROM accompanying this volume]). The system consists of several large belts: Hautavaara, Kindasovo-Man’ga, Koikary, Palalamba, Semch-Suna, Sovdozero, Oster, Bergaul, and a number of smaller belts (Glebovitsky, 2005). According to another version (Svetov, 2005), a single Vedlozero-Segozero greenstone belt is divided into fragments by erosion. According to characteristics given by Svetov (2005), the volcanic-sedimentary associations of greenstone belts are subdivided into two groups dated at 3.02–2.91 Ga and 2.90– 2.85 Ga. The more detailed subdivision is related to the recognition of tectonostratigraphic assemblages formed in certain geodynamic settings and separated by nearly conformable tectonic boundaries in the present-day structure.

Hautovaara Greenstone Belt The Hautovaara greenstone belt in the southwestern Vedlozero-Segozero system of belts (Appendix I-4) includes the most complete succession of tectonostratigraphic associations. The NNE-trending belt extends for 100 km and has a maximum width of 10–12 km. The older basalt-andesite-dacite-rhyolite unit at the base of the sequence includes the corresponding volcanic series, the accompanying volcanic-sedimentary and volcanic-chemogenic sequences, and volcanic and terrigenous rocks that complete the succession. The localization of this association is controlled by a chain of central-type paleovolcanic edifices (Fig. 2.5) that were functioning in subaerial and shallow-water environments. The vent facies are rimmed by block agglomerate tuffs pertaining to facies of explosion ejections and agglomerate flows combined with lenticular members of intercalating pillow lavas, clastic lavas, corded pahoehoe, massive and amygdaloidal lavas with hyaloclastites, and block and lapilli tuffs, along with numerous dacitic and andesitic dikes. Pyrite lenses are related to volcanosedimentary and volcanic-terrigenous facies of dacitic and rhyolitic volcanic activity. The total thickness of the association is estimated at 2.5 km. In composition, the volcanic rocks belong to the evolved calc-alkaline series of normal alkalinity with geochemical characteristics typical of the andesite-dacite-rhyolite series in Phanerozoic island-arc systems. In certain cases, subvolcanic facies are close in composition to adakites (Svetov, 2005). The older komatiite-basalt (komatiite-tholeiite) association, up to 2.7 km in thickness, occurs in the Hautavaara, Koikary, Palaselga (Palalamba), and Sovdozero belts (Svetov, 2005). This is a stratified sequence composed of basalt, basaltic komatiite, and less abundant komatiite with accompanying tuffs, tuffites, and sedimentary exhalative (SedEx) deposits metamorphosed under conditions of greenschist and amphibolite facies. The volume of pyroclastic interlayers is less than 5%. Basalts correspond to quartz, plagioclase-clinopyroxene, and olivine tholeiites, which make up flows of massive or pillow lavas. Tuffs occur as thin interlayers between lava flows, often in association with volcanosedimentary rocks transformed into magnetite-amphibole and carbonaceous schists, as well as chemogenic aluminosilicate rocks. Graywacke and conglomerate of the upper sedimentary member intercalate with graphite schists and massive sulfide deposits. The thickness of volcanosedimentary rocks does not exceed 40–60 m (Rybakov and Golubev, 1999). Basaltic komatiite is predominant among high-Mg volcanic rocks. Sporadic komatiites are represented by Al-undepleted variety. Spinifex structure in the layered komatiite and basaltic komatiite flows is confined to their central or upper parts. The thickness of flows


34

Chapter 2

Figure 2.5. Schematic geological map of the Vedlozero-Segozero system of greenstone belts, modified after Svetov (2005). (1, 2) Paleoproterozoic: (1) rapakivi granite, (2) volcanic-sedimentary complexes; (3–11) Neoarchean: (3) sanukitoid-type diorite and granodiorite, 2.74 Ga; (4) granite, 2.87–2.85 Ga; (5) gabbrodiorite, (6) gabbronorite, (7) mafic-ultramafic complex, (8) highMg gabbro; (9–11) volcanic-sedimentary complexes of greenstone belts: (9) andesitic dacite metavolcanic rocks, including adakite-type varieties and related metasedimentary rocks, 2.86– 2.85 Ga; (10) komatiite-basalt association, 2.95–3.0 Ga; (11) calc-alkaline and adakite metavolcanic rocks, 2.86– 2.85 Ga; (12) Mesoarchean amphibolites, 3.15–3.0 Ga; (13) Neoarchean and Mesoarchean granite gneisses; (14) paleovolcanic centers (numerals in figure) of (1–5) Hautovaara and (6–9) Semch-Koikary greenstone belts: 1—Nälmozero, 2—Ignoila, 3— Hautavaara, 4—Maselga, 5—Chalka, 6—Janish, 7—Korbozero, 8—Elmus, 9—Semch, (15) master faults.


Mesoarchean Kola-Karelia continent varies from 0.5–1.0 to 50–70 m. The komatiitic sections are distinguished by unusually high volumes of pyroclastic material, occasionally reaching 10%–15%. This apparently implies that basaltic komatiites were gas and water saturated and atypical in this respect of deep mantle-plume komatiites (Arndt et al., 1997). In addition to komatiite and tholeiite lavas, magnesian gabbro and completely serpentinized ultramafic rocks vague in nature are noted in the sections. In general, the komatiite-tholeiite association of the Vedlozero-Segozero belt is interpreted as a product of backarc magmatism with probable participation of a mantleplume component (Svetov, 2005). The younger andesite-dacite association consists of intermediate and felsic volcanics. Pyroclastic rocks prevail over lavas. The most complete section has been described in the Koikary belt. The thoroughly studied Janish paleovolcanic edifice is composed of lava breccia, lava, and block agglomerate tuff; the conduit is filled with subvolcanic dacite; chemogenic sediments were deposited in the crater lake. Tuff, tuffite, tuffstone, tuffaceous conglomerate, and sedimentary rocks occur at the periphery of the volcano. Subvolcanic bodies are composed of dacite and rhyolite. The total thickness of the volcanic association is 900 m. Andesitic dacite is close in geochemistry to volcanic rocks of the Andean-type active continental margins. Two rock associations of limited abundance have been revealed in the upper part of section in the Hautovaara belt. The basaltic association, 600 m in thickness, is composed of pillow and massive basalts with sporadic lenses of hyaloclastites and tuffs, and graywacke and conglobreccia interlayers. The upper sedimentary association, ~200 m in apparent thickness, consists of carbonaceous schist, dacitic tuff, tuffite, and chert (Svetov, 2005). The belts of the Vedlozero-Segozero system host pyrite ore bodies and small occurrences of Cu-Ni–sulfide mineralization, probably related to komatiitic volcanism, as well as small iron occurrences, spodumene pegmatites, Cu-Mo porphyry, and Au mineralization (Rybakov, Golubev, 1999). South Vygozero Greenstone Belt The South Vygozero greenstone belt occupies a somewhat isolated position. The fragmentary exposures predetermined the subdivision of this belt into separate areas or structures with outcrops of greenstone rocks pertaining to the komatiite-tholeiite and calc-alkaline series. The komatiite-tholeiite association is dated with the Sm-Nd method at 3.05–2.91 Ga (Table 2.5). The positive initial εNd = +1.6 indicates that the komatiite-tholeiite association was formed by melting of depleted mantle without contamination with material from the ancient continental crust. 2.2.2. Sumozero-Kenozero System of Greenstone Belts The greenstone belts of the Sumozero-Kenozero system extend along the northeastern boundary of the Vodlozero microcontinent (Appendix I-4 [on CD-ROM accompanying this volume]) for ~300 km at a width of 50 km (Kulikov and Kulikova, 1979). The system consists of a series of conjugate synforms and

35

homoclinal structural elements composed of sedimentary and volcanic rocks framed by synkinematic TTG granitoids. The area remains poorly studied because of insufficient exposures, so that most information is taken from geophysical surveying and drilling results. The Kamennoozero, Senegozero, Toksha-Kenozero, Volotsk, and Matkalahti belts are delineated most distinctly. Kamennozero Greenstone Belt According to the results of geological mapping (data of V.M. Tytyk cited after Kuleshevich et al., 2005), the central area of the Kamennozero greenstone belt is a gentle synform, V-shaped in plan view and 40 × 20 km in size (Appendix I-4 [on CD-ROM accompanying this volume]; Fig. 2.6). It is accompanied by complexly built folded branches that extend along the NW-trending faults. It is suggested that folding was related to emplacement of nearly conformable lenticular intrusive bodies (diorite, tonalite, granodiorite, trondhjemite) accompanied by rhyodacitic dike swarms. The total thickness of the section is approximately estimated at 5000 m (Kulikov and Kulikova, 1979). The volcanic and sedimentary rocks often retain their primary structure and texture, which make it possible to restore their original nature and facies. The greenstone belts in the southern part of the area directly border on the granite-gneiss complex of the Vodlozero microcontinent. Intense tectonization of rocks and development of thick (up to 100 m) zones of mylonitization and foliation plunging to the northeast indicate the tectonic character of the boundary. On the basis of regional relationships (Appendix I-1), it can be concluded that the initial thrust-nappe structure originated in the Archean. In the present-day structure, the Paleoproterozoic reverse and thrust faults also play important roles and facilitate rejuvenation of the Archean surfaces of tectonic transport. The sedimentary-volcanic sequence along of the northern boundaries of belt is in contact with a plagiogranitic pluton that incorporates small bodies of potassic granites. Numerous roof pendants and amphibolite xenoliths indicate the intrusive nature of the aforementioned pluton. The relationships of the greenstone section with granitoids of the northern framework most likely are tectonic, as is shown in Figure 2.6 and Appendix I-1. The volcanic-sedimentary section of the Kamennozero belt consists of the lower volcanic-sedimentary, the middle volcanic, and the upper volcanic-sedimentary sequences (Kulikov and Kulikova, 1979; Sokolov, 1984). A certain overturn in the stratigraphy had been done by Puchtel et al. (1999), who subdivided the section of the Kamennozero belt in the older komatiitetholeiite and the younger calc-alkaline associations. The latter one combines subvolcanic, volcanic, and tuffaceous sedimentary rocks of the basalt-andesite-dacite-rhyolite series with a predominance of dacite and rhyolite. According to the recent detailed study by Kuleshevich et al. (2005), the sections of the Kamennozero belt and other belts of the Sumozero-Kenozero consist of the lower Kumbuksa-Savino and the upper Idel formations. The lower formation is composed of metatholeiites, carbonaceous schist, SedEx quartzite, carbonate-quartz-sericite schist after tuff and tuffite, komatiite,


36

Chapter 2

TABLE 2.5. AGE DETERMINATIONS ON THE MESOARCHEAN VOLCANIC ROCKS OF THE VEDLOZERO-SEGOZERO AND SUMOZERO-KENOZERO SYSTEMS OF GREENSTONE BELTS No. Complex, group, Rock, process Age (Ga) Method References massif Vedlozero-Segozero system of greenstone belts Hautovaara greenstone belt The older basalt-andesite-dacite-rhyolite (BADR) association 1 Ignoil neck Andesitic dacite 3.00 U-Pb thermal Sergeev (1989) ionization mass 2 Oster structure Subvolcanic andesite 3.02 Svetov (2005) spectrometry (TIMS) 3 Palaselga structure Andesitic dike 3.00 Lobikov (1982) zircon age 4 Andesitic lava 2.95 Ovchinnikova et al. (1994) 5 Volcanic rocks BADR association 2.92 Whole-rock Sm-Nd Svetov and Huhma (1999) isochron The older komatiite-basalt (komatiite-tholeiite) association 6 Subvolcanic rocks Dacite dike 2.94 U-Pb TIMS zircon Bibikova and Krylov (1983) age 7 2.86 Samsonov et al. (1996) 8 Gabbrodiorite intrusion 2.89 Sergeev et al. (1983) 9 Volcanic rocks 2.92 Whole-rock Sm-Nd Svetov et al. (2001) isochron The younger andesite-dacite association 10 Hautovaara pluton Granodiorite 2.85 U-Pb TIMS zircon Tugariniov and Bibikova (1980) age 11 R hyolite-dacite dike 2.85 Ovchinnikova et al. (1994) 12 Yanish edifice Felsic volcanics 2.86 Samsonov et al. (1996) South Vygozero greenstone belt 13 Komatiite-tholeiite Volcanic rocks 2.91 Whole-rock Sm-Nd Lobach-Zhuchenko et al. (1999) association isochron 14 3.05 Samsonov et al. (1996) Sumozero-Kenozero system of greenstone belts Kamennozero greenstone belt 15 Rocks of the komatiite- Volcanic rocks 2.9 ± 0.1 Whole-rock Sm-Nd Puchtel et al. (1999) tholeiite and BADR isochron associations 16 Komatiite-tholeiite Volcanic rocks 2.9 ± 0.1 Whole-rock Pb-Pb association isochron 17 BAD R association Rhyolite 2.875– U-Pb TIMS zircon 2.873 age 18 Subvolcanic rocks Rhyodacite 2.876 U-Pb TIMS zircon Bibikova after age Puchtel et al. (1999) Volotsk greenstone belt 19 Komatiite-tholeiite Volcanic rocks 2.85 Whole-rock Pb-Pb Puchtel et al. (2007) association isochron 20 2.9 Whole-rock Pb-Pb isochron 21 2.88 Whole-rock Re-Os isochron Postgreenstone gabbroic and granitic intrusions 22 Oster intrusion Granite 2.88 U-Pb TIM S zircon Kovalenko and Rizvanova age (2000) 23 Shilos intrusion Tonalite 2.85 Lobach-Zhuchenko et al. (1999) 24 Semchen intrusion Gabbrodiorite 2.85 Sergeev et al. (1983) 25 Small body in the Leucogabbro 2.84 Lobach-Zhuchenko and Palalamba belt Levchenkov (1985)

basaltic komatiite, and tholeiitic basalt. The upper part of the section is made up of sericite-quartz, carbonate-chlorite-sericitequartz, and carbonaceous schists developed after tuffaceous sedimentary rocks corresponding in composition to intermediate and felsic volcanics, as well as pyrite lenses and interbeds, basalt, basaltic andesite, and lavas, tuffs, and carbonaceous schists. Mafic rocks (amygdaloidal basalts, basaltic andesite, agglomerate tuff) play an important role in the upper sequence, which fills the synform core. It is possible that this stratigraphic chart hides a section of the upper formation doubled by folding and thrusting in the eastern limb of Kamennozero synform. The deposition of the sedimentary-volcanic sequence was followed by emplacement of a series of nearly conformable len-

ticular TTG and granite intrusive bodies accompanied by rhyodacite dike swarms. Prior to discussion of the geochemical and isotopic data, mostly published by Puchtel et al. (1999), an important remark should be made. The komatiite-tholeiite association proper is known only from the lowermost part of the section (Puchtel et al., 1999, their figure 2). Considerable numbers of tholeiite samples, in turn, have been taken up section from the middle or upper parts of the volcanic-sedimentary sequence, in accord with other authors (Rybakov and Golubev, 1999; Kuleshevich et al., 2005). In other words, it may be suggested that oceanic tholeiites intercalate with basalt-andesite-dacite-rhyolite volcanics in the upper part of the late calc-alkaline association after Puchtel et al.



Figure 2.6. Schematic geological map of the Kamennozero greenstone belt, after Kuleshevich et al. (2005). (1) Early Paleoproterozoic (Sumian): volcanic-sedimentary complexes of the Vetreny belt, unspecified; (2–9) Archean: (2–6) intrusive rocks: (2) subvolcanic adakite-type rhyodacite, (3) granite, (4) tonalite-trondhjemite-granodiorite (TTG); (5) gabbroic rocks, (6) metaperidotites (serpentine-magnetite-talccarbonate rocks); (7–9) greenstone volcanic-sedimentary lithotectonic associations: (7) upper sedimentary–volcanic basalt–andesite–dacite– rhyolite lithotectonic association, including tholeiites of volcanic plateaus, after Puchtel et al. (1999); (8) middle sedimentary-volcanic basalt-andesite-dacite-rhyolite lithotectonic association; (9) lower komatiite-tholeiite lithotectonic association, a fragment of volcanic plateau, after Puchtel et al. (1999); (10) granite gneisses, unspecified; (11) faults, mainly reverse-thrust; (12) ore occurrences mentioned in the text (numerals in figure): 1—North Vozhma, 2—Upper Vozhma, 3—Lebyazhinka, 4–6—Zolotye Porogi, 7—West Svetlozero.

(1999). Another explanation assumes that the lower komatiitetholeiite association emerged at the surface in the middle part of a volcanic-sedimentary part of the section owing to folding, as seen in Figure 2.6. Thus, the section at the southeastern closure and the eastern limb of the V-shaped synform in the Kamennozero belt can be summarized as a succession of the following tectonostratigraphic units (from bottom up): the lower komatiite-tholeiite unit → middle volcanic-sedimentary basalt-andesite-daciterhyolite unit → upper unit formed by alternation of the volcanicsedimentary basalt-andesite-dacite-rhyolite association presumably with incorporation of oceanic tholeiites of the same type as participating in the komatiite-tholeiite association → subvolcanic bodies of adakitic rhyolite localized in the upper part of the

37

section (Puchtel et al., 1999, their figure 2), or conversely, in its lower part (Fig. 2.6). A series of metaperidotite sills transformed into serpentinemagnetite-talc-carbonate rocks is localized along the boundary between the middle and the upper associations. The total extent of these sills attains 20 km. Pillow, variolitic, and massive tholeiitic basalts dominate in the lower komatiite-tholeiite association (Fig. 2.6). Hyaloclastic tuff and breccia are of subordinate abundance as thin interlayers among lavas. In the southern part of the belt, tholeiitic basalts are associated with komatiites and basaltic komatiites. Carbonaceous schist, quartzite, and carbonate-quartz-sericite schists after tuff and tuffite also play a substantial role. Komatiites belong to the Al-undepleted type with flat REE patterns and low positive Nb anomalies (Nb/Nb* = 1.2). Variations in chemical composition of komatiites are readily explained by olivine fractionation. The isotopic parameters of komatiite-tholeiite association show that it should be regarded as a pure product of mantle melting without any involvement from the ancient sialic crust (Puchtel et al., 1999). The middle sedimentary-volcanic basalt-andesite-daciterhyolite association of the Kamennozero belt is localized in the eastern part of the V-shaped synform. The volcanic section, within an extended zone 4–6 km wide, is composed of daciterhyolite tuff and tuffite with interlayers and lenses of carbonaceous slate, phyllite, and metamorphosed SedEx silicic rocks (Popov et al., 1986; Sokolov, 1981). Thin basaltic and andesitic lava flows alternate with pyroclastic beds. The dominating felsic volcanics have lenticular banding structure, and as far as can be judged from metamorphosed and deformed rocks, they originally were pyroclastic breccias and crystal clastic tuffs making up thick (up to 100 m) members. Massive units of felsic volcanics are rare. The basic, intermediate, and felsic volcanic rocks belong to the calc-alkaline basalt-andesite-dacite-rhyolite series characteristic of island arcs. The calc-alkaline basalts have moderately fractionated REE patterns with small Eu (Eu/Eu* = 0.88) anomalies and minimums of high field strength element (HFSE) incompatible elements (Nb, Zr, Ti). The felsic tuff fits ultrasilicic (77.7–81.1 wt% SiO2) and low-Al rhyolite. Geochemical simulation shows that rhyolite of the Kamennozero belt could have been a product of lowpressure fractionation of basalts and subsequent differentiation of andesitic magma (Puchtel et al., 1999). The upper mixed association apparently is an alternation of the sedimentary-volcanic basalt-andesite-dacite-rhyolite association and tholeiitic basalts of the same type as those characteristic of the lower komatiite-tholeiite association. The subvolcanic rhyodacite bodies in the northern part of the belt vary in thickness from decimeters to a few meters and occur among tholeiitic pillow basalts (Fig. 2.6). In composition, these rocks have no analogs in the lower part of the section and belong to the adakitic series. The geochemical simulation shows that rhyolite from the northern Kamennozero belt could have been a product of melting of amphibolite, similar in geochemistry to


38

Chapter 2

metabasalt of the early komatiite-tholeiite association (Puchtel et al., 1999; Bogatikov et al., 2006). The concordant Sm-Nd and Pb/Pb ages of rocks pertaining to the komatiite-tholeiite association of the Kamennozero structure are 2.9 ± 0.1 Ga (Table 2.5). The initial positive εNd value (2.7 ± 0.3) indicates a juvenile origin of magma. The U-Pb zircon age of rhyolites from the basalt-andesite-dacite-rhyolite series in the middle rock association is estimated at 2875–2873 Ma. These estimates coincide within uncertainty limits with the age of the komatiite-tholeiite association. The Sm-Nd isotopic parameters of basalt, andesite, and rhyolite show that these rocks are genetically interrelated. The positive εNd value (2.1–3.1) indicates that these rocks were derived from depleted magma sources. Subvolcanic rhyodacite bodies in the northern Kamennozero belt were formed ca. 2876 ± 5 Ma (U-Pb zircon age; oral communication by E.V. Bibikova cited after Puchtel et al., 1999), synchronously with deposition of the middle sequence. The initial positive εNd value (2.5–4.5) indicates a lack of ancient sialic contaminant (Puchtel et al., 1999). The chemical composition of komatiites from the lower rock association corresponds to a high degree of partial melting of the depleted mantle source. The MgO content in the spinifex zone of komatiites reaches 30 wt% and gives evidence for a high liquidus temperature (~1570 °C). Taking into account the Al-undepleted composition of komatiite from the Kamennozero belt, the melting depth should be accepted beyond the field of majorite stability, i.e., at 7 kbar and ~700 °C), which are typical of metamorphic sole of many Phanerozoic ophiolitic nappes (Shchipansky, 2008). The age of the Iringora structure is characterized by one determination. The U-Pb zircon age of a dacitic sill in the islandarc complex is 2.78 ± 0.01 Ga (Bibikova et al., 2003; Shchipansky et al., 2004; Table 2.9). Kalikorvi Greenstone Structure The Kalikorvi greenstone structure (Appendices I-1 and I-4 [on CD-ROM accompanying this volume and at end of volume]) is a complexly built synform, ~8 × 8 km2 in area, which is situated in the boundary zone of the Kovdozero microcontinent (Mil’kevich et al., 2007). The lower sequence consists of garnet and garnet-clinopyroxene rocks, which correspond to tholeiitic


54

Chapter 2

basalts in bulk chemical composition. Metapyroxenite and metaperidotite lenses are incorporated into amphibolites. Sporadic interlayers are composed of biotite-amphibole and garnet-bearing gneisses (meta-andesites and metadacites), as well as kyanitebearing gneisses (metagraywackes). Sporadic thin interlayers resemble basaltic komatiite and Fe-Ti tholeiite in chemical composition. The upper sequence is made up largely of garnet-biotite and partly kyanite-bearing gneisses (metagraywackes); members of fine intercalation of various rocks, including pyroxene-garnetplagioclase amphibolites (metatholeiites with meter-thick interlayers of basaltic komatiite), biotite-plagioclase amphibolites partly with garnet and clinopyroxene (subalkali metabasalts and metapicrobasalts), and biotite-plagioclase gneisses partly with garnet (metadacites, meta-adakites). Primary textures and contacts have not been retained. The age of rocks pertaining to the Kalikorvi greenstone structure are bracketed in the narrow time interval between 2.79 and 2.77 Ga (Table 2.9). Tunguda, Pebozero, Parandovo-Nadvoitsk, and East Idel Greenstone Belts The Tunguda, Pebozero, Parandovo-Nadvoitsk, and East Idel greenstone belts in the southern part of the ParandovoTiksheozero system of greenstone belts (Appendix I-4 [on CDROM accompanying this volume]) are studied less completely as compared with greenstone structures of the North Karelian belt. Granitoid rocks in this part of the Kovdozero microcontinent pertain to the tonalite-trondhjemite and gabbro-dioritegranodiorite-granite series. No geological evidence is known to validate the existence of pregreenstone granite gneisses in the Kovdozero microcontinent. The greenstone section of the Pebozero belt is composed of amphibolites, amphibole gneisses, mica-, amphibole-, and garnet-bearing staurolite-kyanite schists and gneisses (metavolcanic rocks of basaltic, andesitic, dacitic, and less frequent rhyolitic compositions). Metaterrigenous rocks (mainly metagraywackes) occupy 10%–40% of the bulk volume; carbonaceous and carbonate schists occur in insignificant amounts; interbeds of polymictic conglomerate and quartzitic rock are noted. The total thickness of this section is as large as 5000 m. The age of rocks in the Pebozero succession is estimated at 2.81–2.79 Ga (Table 2.9). Geochronology and Geodynamic Interpretation of Mesoarchean History of the Kovdozero Microcontinent Turning to the geodynamic settings in the history of the Kovdozero microcontinent, let us summarize its main features. (1) The Parandovo-Tiksheozero island-arc paleosystem was formed during a short time interval ca. 2.80–2.78 Ga (probably 2.81–2.77 Ga), i.e., over 20–40 m.y. (2) Volcanic and sedimentary rocks that were formed in a spreading ridge in the Archean ocean, subducting island arc, zone of forearc spreading, and in a deep-water trench took part in the structure of the island-arc paleosystem. In other words, an almost

complete set of rock associations interacting in the continentocean zone of convergence participated in the structure of the Parandovo-Tiksheozero island-arc paleosystem. (3) Fragments of one of the world’s oldest (ca. 2.8 Ga) suprasuduction ophiolitic associations known to date have been retained in the Iringora structure of the North Karelian greenstone belt. As a result of subsequent structural transformation and metamorphism, the primary pseudostratification of the ophiolitic complex has been destroyed to such a degree that it is impossible to reliably determine its thickness. Extremely important information has been retained in geochemical signatures. The Neoarchean boninite series is very close in geochemistry to the well-known pillow lavas from reference Troodos ophiolites used for recognition of high-Ca boninite series (Crawford et al., 1989). Surprisingly similar geochemistry and isotopic composition of Neoarchean and late Mesozoic boninite series can be explained only by their similar formation conditions (Shchipansky et al., 2004). The geology of the Iringora structure shows that the ophiolitic complex has been thrust over rocks of the mature island arc and volcanic-sedimentary wedge. This implies that ophiolites were formed in a setting of forearc spreading, as this is accepted for most Phanerozoic and recent analogs (Beccaluva and Serri, 1988; Pearce et al., 1992; Stern and Bloomer, 1992). (4) As in the case of the Khetolambina microcontinent, the successions of greenstone belts pertaining to the ParandovoTiksheozero system are characterized by participation of amphibolites enriched in Fe-Ti oxides (Fe-Ti gabbro and probably volcanic rocks). It is known that Fe-Ti gabbro is one of the important components of the third layer of the oceanic crust (layered gabbro) in slow-spreading ridges (Dick et al., 1999, 2000; Blackman et al., 1998; Grimes et al., 2008). This correlation is considered in detail later herein (section 2.4). (5) The available data do not contain evidence of an ancient pre–island-arc or pregreenstone basement of the Kovdozero microcontinent. The plutonic complexes could have been formed contemporaneously with accumulation of volcanic-sedimentary associations related to them genetically. (6) The grade of metamorphism was approximately similar in various fragments of the island-arc paleosystem accessible for observation. The P-T parameters of peak metamorphism M1 in the Hizovaara and Iringora structures (670–730 °C and ~7.6 kbar) correspond to the high-temperature boundary between amphibolite and pyroxene-plagioclase granulite facies. The pressure value makes it possible to estimate the depth of metamorphism at 24–29 km. Further interpretation of these parameters depends on the accepted age of metamorphism. In general terms, high-P and high-T metamorphism could have been related to (1) evolution of the Parandovo-Tiksheozero island-arc paleosystem ca. 2.78 Ga, as suggested by Shchipansky (2008); (2) plume activity, extension, and granulite-facies metamorphism, which gave rise, in particular, to formation of the Chupa granulite-gneiss belt, as well as the Notozero complex of granulites and enderbites (chapter 3),


Mesoarchean Kola-Karelia continent or (3) emplacement of the Topozero charnockite in the early Paleoproterozoic (section 8.1.1). (a) The first version allows us to put forward attractive inferences. With allowance for transitional crust of the mature island arc, a depth of 24–29 km corresponds to the lower part of the crust. Thus, the assembly of imbricated tectonic sheets can be regarded as a sort of island-arc crust. This suggestion is consistent with the fact that there are no relict volcanic centers, characteristic, for example, of the Vedlozero-Segozero island-arc paleosystem, where metamorphic grade does not exceed the level of epidote-amphibolite facies. (b) The second and third versions (age of metamorphism is 2.75–2.72 Ga or early Paleoproterozoic) assume superposed character of metamorphism, which rules out this phenomenon from the Mesoarchean history of the Kovdozero microcontinent proper. Since there is no information on age of metamorphism, the question has no answer at present. (7) According to Miller et al. (2005), the Kovdozero microcontinent is a tectonic sheet, ~3 km thick, in the section along seismic profile 4B. The sheet is deformed into a synform (lopolith-like body). The boundaries of this microcontinent shown in geological maps give evidence for thrusting of its marginal zones over the framework. In combination with the estimated depth of metamorphism, these features of the Kovdozero microcontinent can be interpreted as evidence for squeezing out under conditions of collisional compression accompanied by rapid growth of the orogen and its denudation. In particular, such a model explains exposure of the Neoarchean subduction zone at the present-day surface. It should be kept in mind that in the proposed reconstruction, we have to omit a question on the age of the collisional event (whether it is Neoarchean or Paleoproterozoic). Since vigorous and diverse manifestations of Paleoproterozoic magmatism, sedimentation, and tectogenesis are related to the boundary zone between the Belomorian province and the Kare-

55

lian craton, it is most probable that tectonic squeezing of the Kovdozero microcontinent was an integral effect of consecutive collisional events in the Neoarchean and Paleoproterozoic; however, no specific evidence for such a sequence of events is available to date. Thus, the Kovdozero microcontinent can be interpreted as a mature, and moreover, very rapidly matured island arc. The arc could have originated in an intra-oceanic setting or it could have been a broken-off margin of the Karelian craton or the Khetolambina microcontinent. The arc was rapidly formed between the Kuhmo-Segozero and Khetolambina microcontinents. As a result of their collision and probably owing to subsequent tectonic processes, the arc was squeezed out, and its marginal segments thrust over the framework. The main events in the geological history of the Kovdozero microcontinent apparently developed in the following succession: growth of a spreading ridge, probably predated by separation of the eastern margin of the Karelian craton or western margin of the Khetolambina microcontinent → subduction, most likely beneath the Khetolambina microcontinent, with formation and maturing of the island arc ca. 2.80–2.78 Ga → forearc spreading and formation of boninite on the oceanic plate → accretion– collision → high-temperature metamorphism (ca. 2.78, 2.73– 2.71, or ca. 2.45 Ga) → collisional compression and squeezing of the granite-greenstone complex over the framework. The final Neoarchean events were accompanied by formation of late conglomerate, quartz arenites, and coarse tuffaceous conglomerate, which were involved in further tectonic displacements. Summing up the features of structure, composition, and metamorphism of rocks in the Kovdozero microcontinent, we reach the conclusion that roots of the island-arc paleosystem are exposed at the present-day erosion level, i.e., a zone of accretion of subducting material to the base of the arc and underplating of the crust. In our opinion, this suggestion gives a new incentive to study the Kovdozero microcontinent.


56

Chapter 2

2.4. MESOARCHEAN–NEOARCHEAN BELOMORIAN ECLOGITE PROVINCE Ksenia A. Dokukina, Alexander N. Konilov, Michael V. Mints, Elena A. Belousova, Peter A. Dokukin, Tatiana V. Kaulina, Lev M. Natapov, and Konstantin V. Van

In this section, we present a review of the recently discovered Archean Belomorian eclogite province in the eastern part of the Fennoscandian Shield (Fig. 2.9). The tectonic significance and specific signatures of the Archean eclogites, as well as their implications for timing of the onset of modern-style plate tectonics, are considered further in chapter 15. For many years, it was assumed that an eclogite-facies metamorphic regime could not have existed during early stages of Earth’s history due to the higher geothermal gradients (Green, 1975; Baer, 1977). However, Mesoarchean eclogite xenoliths have been found in kimberlite pipes from the Kaapvaal and Siberia cratons, where the oldest diamonds containing inclusions of eclogite-facies omphacite and garnet have been dated at ca. 2.9 Ga (Gurney et al., 2010, and references therein). Paleoproterozoic eclogites were recognized for the first time in the northwestern Scotland Highlands in 1984 (Sanders et al., 1984; Sanders, 1988) within the Glenelg-Attadale basement inlier of the Caledonian orogen. Then, Proterozoic eclogites were found in many other localities. The early Precambrian eclogites from the eastern Fennoscandian Shield were first described by Sudovikov (1936), Batieva (1958), and Smirnova and Baboshin (1967); however, their petrological and isotopic characteristics attracted attention only recently. The Archean eclogites were first discovered by V. Barzhitsky in 2002 on a roadside outcrop near the Salma Strait of Lake Imandra in the Kola Peninsula (Shchipansky et al., 2005). The first detailed petrogenetic study and an attempt to date the Belomorian eclogites were reported by Volodichev et al. (2004), who described the minor eclogitic bodies that occur within a tectonic mélange zone near the Karelian village of Gridino at the western shore of the White Sea. These authors concluded that the origin of the newly discovered eclogites was caused by Archean subduction (Volodichev et al., 2004, 2005). The aforementioned papers gave impetus to numerous detailed studies of eclogite bodies (Volodichev et al., 2004, 2005, 2008, 2009; Dokukina et al., 2005, 2009, 2010, 2011b, 2014; Dokukina and Konilov, 2011; Aranovich and Kozlovskii, 2009; Kozlovsky and Aranovich, 2008, 2010; Konilov et al., 2011; Mints et al., 2010a, 2010b, 2014b; Mints and Konilov, 2011; Perchuk and Morgunova, 2014; Rosen et al., 2008; Skublov et al., 2010a, 2010b, 2011a, 2011b; Slabunov et al., 2006b, 2006d, 2011b; Travin and Kozlova, 2005; Kaulina, 2010; Shchipansky et al., 2012a, 2012b). The investigations carried out by these research groups have established the main characteristics of eclogite bodies; highlighted the main features of magmatic, metamorphic, and tectonic evolution of eclogites and eclogite-bearing rock associations; and reconstructed the main features of their geodynamic evolution. The interpre-

tation of the geochronological results, the nature and sequence of metamorphism, and reconstructions of geodynamic settings proposed by many authors turned out significantly and, in some cases, quite different. The aim of this section is to outline and examine the most important characteristics of the Belomorian eclogite province. We will review the regional tectonic setting and geological features of the Belomorian eclogite province, the sequence of major magmatic and metamorphic events and corresponding geodynamic settings, and the attempts to reconstruct the history of the origin and transformation of various rock types found within the Belomorian eclogite province. We also discuss the reasons behind different interpretations of the nature and evolution of the Belomorian eclogite province and state our viewpoint. 2.4.1. Geological Background The Archean Karelia and Kola continents (consisting predominantly of granite-greenstone terrains) and the Belomorian accretionary-collision orogen are the major tectonic units of the eastern Fennoscandian Shield (Fig. 2.9). The Belomorian tectonic province has been of particular interest to geologists for a long time (Slabunov et al., 2006a, and references therein). This province was distinguished by repeated episodes of intense deformation and high- to moderate-pressure metamorphism during both the Archean and Paleoproterozoic. Several researchers have suggested that the Belomorian tectonic province is a long-lived mobile belt that was formed along the eastern margin of the Karelia continent as a result of consecutive tectonic events: westward subduction (in present-day coordinates) of the Archean oceanic lithosphere beneath the Karelian continent, accretion of island-arc complexes to the Karelian margin, and final collision ca. 2.75–2.65 Ga (Bibikova et al., 1999; Miller et al., 2005; Slabunov et al., 2006b; Slabunov, 2008). This model is, however, geometrically inconsistent with the crustal-scale seismic-reflection images along the 1-EU seismic geotraverse and 4B cross-traverse, which crossed the Belomorian province (Mints et al., 2009, 2014b). In our understanding, the Mesoarchean and Neoarchean complexes of the Belomorian tectonic province (in traditional meaning) taken together represent an accretionary-collisional orogen formed between 2.88 Ga and 2.76 Ga as a result of eastward subduction beneath the Kola continent. The mafic-ultramafic Central Belomorian greenstone belt separating the Keret and Khetolambina units has been dated at 2.88–2.85 Ga. The available geological, isotopic, and geochemical data on the mafic-ultramafic rocks of the greenstone complex



57

Figure 2.9. Simplified geological maps of the northeastern Fennoscandian Shield, showing the tectonic position of the Belomorian eclogite province, after Mints et al. (2010a) and (2014b): Mints, M.V., Dokukina, K.A., and Konilov, A.N., 2014, The Meso-Neoarchaean Belomorian eclogite province: Tectonic position and geodynamic evolution: Gondwana Research, v. 25, p. 561–584, doi:10.1016/j.gr.2012.11.010. © 2014, reprinted with permission of Elsevier. (Figures 2.10, 2.11, 2.12, 2.13, 2.14, 2.15, 2.17, 2.23, 2.24, 2.26, and 2.31 are also taken from that article.). 1–6 denote eclogite body occurrences of the Salma association: 1—Uzkaya Salma, 2—Shirokaya Salma, 3—Upolaksha, 4—Chalma (Kuru-Vaara quarry), 5—Hangaz-Varaka, 6—Gridino; 7–8—eclogitized dike occurrences of the Gridino association: 7—Krasnaya Guba, 8—Gridino; 9—eclogitized dikes on the islands in Onega Bay, to the south of the area shown on the map.

are compatible with its interpretation as tectonically disrupted and metamorphosed remnants of a Mesoarchean ophiolitic association (Slabunov et al., 2006a; see section 2.3.1). In turn, the Keret nappe, sandwiched between Khetolambina and Inari-Kola microcontinents, contains TTG gneisses dated at 3.00–2.70 Ga and greenstones, as well as numerous eclogite bodies (Mints et al., 2014b; Fig. 2.10).

The overall structural setting combined with the aforementioned data and evidence for emplacement of eclogite bodies into TTG gneisses allow us to interpret the Central Belomorian belt as a Mesoarchean collisional suture zone. The seismicreflection data indicate that the Khetolambina microcontinent continues at a depth to the northeast beneath the Inari-Kola microcontinent, although the boundary zone between these


58

Chapter 2

Figure 2.10. Geological cross section along the 1-EU reflection seismic profile, after Mints et al. (2014b). Reprinted with permission of Elsevier. See Figure 2.9 for location and legend.

Figure 2.11. Location of the Uzkaya Salma eclogite assemblage: outcrop along the northeastern roadside along the St. Petersburg–Murmansk highway, compiled by K.A. Dokukina, N.E. Kozlova, and O.N. Platonova, photo by K.A. Dokukina. Figure is after Mints et al. (2014b); reprinted with permission of Elsevier.


Mesoarchean Kola-Karelia continent two units, which coincides with the Central Belomorian belt, is not clearly expressed in the seismic image (Mints et al., 2009; Mints, 2014b). Based on geological mapping and the results of seismic-reflection profiling, the Keret nappe, together with the overthrust crustal segment, has been interpreted as an active margin of the Archean Inari-Kola microcontinent (Figs. 2.9 and 2.10). According to our understanding, these three units (the Khetolambina microcontinent, the eclogite-bearing active margin of the Inari-Kola microcontinent, and the Central Belomorian suture) represent a coherent Mesoarchean and Neoarchean tectonic assembly, which we refer to as the Belomorian eclogite province (Mints et al., 2010a, 2010b). In the Keret nappe, two distinct eclogite-bearing rock associations are distinguished. The Salma association is attributed to subduction and high-pressure metamorphism of the oceanic lithosphere (Mints et al., 2010a, 2010b; Konilov et al., 2011), whereas the Gridino association is characterized by injection of mafic dikes subsequently metamorphosed under eclogite-facies conditions (Volodichev et al., 2004, 2005; Mints et al., 2010b; Dokukina and Konilov, 2011; Dokukina et al., 2014). Eclogite bodies belonging to the Salma association are localized in the southern Kola Peninsula at the Uzkaya Salma, Shirokaya Salma, Upolaksha, Chalma (Kuru-Vaara quarry), and Hangaz-Varaka occurrences. In addition, eclogitic boudins of the same type are known close to Gridino Village (Volodichev et al., 2004, 2005; Slabunov et al., 2006a). In contrast, the main and most typical and abundant components of the Gridino eclogite association are partly eclogitized dikes of the Gridino swarm, which are well exposed along the White Sea shoreline and adjacent islands. The age of eclogite-facies metamorphism in the Belomorian province has been hotly debated in recent years, whether it was Archean or early or late Paleoproterozoic. We will return to this problem later herein. Salma Eclogite Association The retrogressed eclogites of the Salma association have been investigated in detail at Uzkaya Salma and Shirokaya Salma (Mints et al., 2010a, 2014b; Konilov et al., 2011), at Chalma (Kuru-Vaara quarry; Shchipansky et al., 2012a, 2012b), and on Stolbikha Island and some other islands in the White Sea in the vicinity of Gridino Village (Volodichev et al., 2004; Fig. 2.9). The best-studied large eclogite bodies are exposed along the southern shore of the Uzkaya Salma Strait in Lake Imandra in the Kola Peninsula. The St. Petersburg–Murmansk Highway passes across the strike of these bodies at a distance of 1192 km from St. Petersburg (Fig. 2.11). A 300–500-m-wide eclogite-bearing association is traced to the northeast for more than 4 km along the shores of the Babinskaya Imandra and Ekostrovskaya Imandra lakes. The host TTG gneisses vary in composition from quartz diorite to trondhjemite and contain a diverse range of mafic eclogites, layers and lenses of Fe-Ti eclogites and high-Mg eclogite-facies rocks (piclogites), layers and lenses of garnetites, and garnet-bearing and garnetfree amphibolites. Banding within the gneisses and the contacts

59

of the eclogite bodies steeply dip toward the north-northwest. The Archean rocks are crosscut by vein- and lenslike bodies of the late Paleoproterozoic granitoids and pegmatites intruded under amphibolite-facies conditions. Most of the eclogite bodies are characterized by advanced retrogression and formation of amphibolites, with only locally retained relict domains of highpressure assemblages. Retrogressed eclogites typically consist of poikilitic garnet porphyroblasts (4–5 mm) in a fine-grained lightgreen matrix of Na-poor clinopyroxene-plagioclase symplectite (after omphacite), and minor amphibole and quartz. Olive-green hornblende, plagioclase, and ilmenite replace clinopyroxeneplagioclase symplectite, while garnet is replaced by kelyphitic rims of plagioclase with hornblende. Mafic eclogites are the predominant type among the eclogite bodies. The marginal zones of the larger eclogite bodies are characterized by garnetbearing or garnet-free amphibolites, while their interiors are also amphibolized in some cases. Amphibole growth is also common in the exocontact zones of the late Paleoproterozoic quartz and pegmatite veins and within local fracture zones. Mafic eclogites are often associated with high-Mg rocks (piclogites). Massive coarse- and medium-grained Fe-Ti eclogites can be readily distinguished owing to their rusty coloration. Garnetites consisting of garnet (up to 90%), plagioclase, pyroxene, and quartz, with conspicuous amounts of rutile, form lenticular bodies or extended layers fragmented into boudins. The occasionally intersecting bands range in thickness from a few centimeters to 1 m or more. At least, in some cases, garnetites are metamorphosed Fe-Ti gabbro with titanomagnetite and ilmenite-magnetite mineralization. It is also possible that some garnetite layers represent metamorphosed veins that originally formed in the gabbroic protoliths due to hydrothermal metasomatism processes on the ocean floor or below it. Mafic and ultramafic eclogite-facies rocks can be assigned to the tholeiitic series. The mafic eclogites resemble low-K gabbronorite or tholeiitic basalts (Fig. 2.12). The high-Mg rocks contain 18% MgO). Protoliths of the Fe-Ti eclogitic rocks, resembling olivine gabbro or troctolite, are significantly enriched in FeO* and TiO2 at 18–21 wt% and 1.5–2.5 wt%, respectively. The trace-element abundances point toward a mantle derivation of the magmatic precursors, but they are lower than those of N-MORB. A comparison of spidergrams of all three Salma eclogite types demonstrates their similarity and coherence (Fig. 2.13). Negative Th anomalies and positive Nb anomalies are general characteristics of the Salma eclogites. Consistent positive U, Zr, and Hf anomalies and negative Ba anomalies also emphasize close relationships between piclogites and Fe-Ti eclogites. In the Th/Yb versus Nb/Yb diagram (Gorton and Schandl, 2000; Fig. 2.14), the Salma eclogites cluster near the mean composition of recent MORB. When plotted on the Nb/Y versus Zr/Y diagram (Fitton et al., 1997), the Salma eclogites fall into the field of the mantle-plume array (Fig. 2.15). These features imply a contribution of a mantle plume to the source of the eclogitic protoliths.


60

Chapter 2

Figure 2.12. AFM (Na2O + K2O [A] – FeO + Fe2O3 [F] – MgO [M]) diagram (Kuno, 1968): petrochemical types of the rocks in the Salma and Gridino eclogite associations and some other Paleoproterozoic and Archean mafic rocks from adjacent areas, after Mints et al. (2014b). Reprinted with permission of Elsevier.

The layering formed by alternating “normal,” high-Mg, and Fe-Ti eclogitic rocks together with their compositional and structural features suggest that the protolith assemblage consisted of intercalating “normal” gabbronorite, Fe-Ti gabbro, and olivine gabbro with local troctolitic rocks resembling the gabbroic suite (seismic layer 3) of the present-day oceanic crust of the slowspreading Southwest Indian Ridge (Dick et al., 2000). Several tens of eclogite blocks are randomly distributed through the TTG gneisses exposed at walls of the KuruVaara quarry (the Chalma eclogite locality). The kyanite- and orthopyroxene-bearing eclogite varieties known in the KuruVaara quarry have not been found at any other locality. Some of eclogite blocks furnish evidence for partial melting and initial segregation of felsic melt before or during eclogite-facies metamorphism as veins and melt percolation channels (Shchipansky et al., 2012a, 2012b). Gridino Eclogitized Dike Swarm Eclogite-bearing mafic dikes of the Gridino swarm are localized along the White Sea shoreline and on the adjacent islands near the village of Gridino (Fig. 2.16). These dikes have been described in detail in a number of previous works (Volodichev et al., 2004, 2005; Sibelev et al., 2004; Travin and Kozlova, 2005; Slabunov et al., 2006a; Stepanova and Stepanov, 2010, and references therein; Dokukina and Konilov, 2011). A tectonic mélange zone, ~50 km long and 10 km wide, is traced along the White Sea coast from northwest to southeast. Field observations and regional geological mapping revealed that the mélange zone is

formed by and associated with a system of tectonically imbricated slices dipping mainly to the northeast and east (Fig. 2.16). The mélange zone consists of migmatized granite gneisses with abundant amphibolite inclusions as well as of diverse orthogneisses and paragneisses with extended lenses of intensely deformed rocks. The dike swarm trending in the near-meridional direction (Fig. 2.16) consists predominantly of gabbronorite and metagabbro dikes. The strain state of dikes varies considerably. Undeformed dikes retain intrusive contacts and discordantly offset banding in the felsic gneisses. Deformed dikes vary in thickness and degree of deformation, including folding, boudinage, and migmatization. Extreme deformation of the dikes led to their breakup into pods and lenses concordant with foliation of host gneisses (Fig. 2.17). The successive metamorphic events that occurred under eclogite-, high-pressure granulite-, and amphibolite-facies conditions are recorded in various dikes, mafic pods, and lenses. Before proceeding to the discussion on the origin and implications of the eclogites, it should be emphasized that beginning from the paper by Stepanov (1981), the Gridino eclogitized dike swarm was considered and treated as a part of the Paleoproterozoic mafic intrusive bodies referred to as drusites (Stepanova and Stepanov, 2010, and references therein; see section 8.1.1 in this volume), despite the obvious difference between eclogitized dikes and drusite intrusions in metamorphic grade and their spatial and structural positions. It should also be noted that the original assignment of the Paleoproterozoic age of the



61

Figure 2.13. Primitive mantle–normalized (Hofmann, 1988) trace-element patterns of eclogites, after Mints et al. (2010b, 2014b). Reprinted with permission of Elsevier. N-MORB—normal mid-ocean-ridge basalt. See text for explanation.


62

Chapter 2

Figure 2.14. Ta/Yb–Th/Yb plot. N-MORB—normal mid-ocean-ridge basalt; E-MORB—enriched mid-ocean-ridge basalt; IOA—intraoceanic arcs; ACM—active continental margins; WPVZ—within-plate volcanic zones; and WPB—within-plate basalts (after Pearce, 1983; revised by Gorton and Schandl, 2000). The Salma and Gridino eclogite association is based on data from Volodichev et al. (2004), Slabunov et al. (2008), Mints et al. (2010b), and Dokukina and Konilov (2011). Figure is after Mints et al. (2010b, 2014b). Reprinted with permission of Elsevier.

dikes was based solely on geological inferences, such as distribution, outer appearance and morphology of intrusive bodies, the relationships with host rocks, and the bulk composition of dikes, although this information alone is not sufficient for correlation between eclogitic dikes and drusites. Therefore, it is necessary to take into account their geochemical characteristics and affinities (Mints et al., 2010b; Dokukina and Konilov, 2011). The metagabbroic and metagabbronoritic dikes and the Gridino eclogite bodies, for which subduction origin is suggested, belong to the tholeiitic series. A distinctive feature of the metagabbronoritic dikes is their enrichment in Mg and Cr (Mg# is 0.65–0.78; Cr content is 940–1960 ppm) combined with low Ti content (0.41–0.85 wt% TiO2). In contrast, many metagabbro bodies are significantly enriched in FeO* (up to 16 wt%) and TiO2 (up to 2.0–2.5 wt%). The chemical compositions of Paleoproterozoic drusite intrusions are also plotted in an AFM (Na2O + K2O [A] – FeO + Fe2O3 [F] – MgO [M]) diagram (Fig. 2.12) for comparative purpose. Their compositions are clustered within two isolated fields that correspond to the gabbro-anorthosite and troctolite-gabbronorite rocks, respectively. The compositional fields of eclogitized dikes from the Gridino swarm and of the Paleoproterozoic drusites differ significantly. The eclogitized dike compositions plot primarily in the fields of intraplate basalts and basalts from active continental margins. In the Nb/Y–Zr/Y diagram (Fig. 2.15), the data points of the dikes are broadly distributed across the fields of both primitive mantle (PRIMA) and continental crust. In the Th/Yb–Ta/Yb diagram (Fig. 2.14), the mafic dikes also show a wide range in composition, overlapping

Figure 2.15. Zr/Y–Nb/Y plot (after Fitton et al., 1997). The Salma and Gridino eclogite association is based on data from Volodichev et al. (2004), Slabunov et al. (2008), Mints et al. (2010b), and Dokukina and Konilov (2011). Primitive mantle and normal mid-oceanridge basalt (N-MORB) values are after Hofmann (1988), and continental crust is after Rudnick and Fountain (1995). Figure in general is after Mints et al. (2010b, 2014b). Reprinted with permission of Elsevier.



63

Figure 2.16. Geological sketch map of the Gridino dike swarm. Location of the image is indicated by asterisks under the numbers 6 and 8 in Figure 2.9, right. Compiled on a basis of geological sketch map of the Gridino Village area in Slabunov et al. (2012, their figure 4).

the fields of the Gridino subduction-related eclogites as well as of volcanics from the active continental margin contaminated with crustal material. Thus, a distinct relationship between the protolith of the subduction-related eclogites and the mafic dikes is deduced from the primary geochemical data and a set of geochemical discriminant diagrams (Mints et al., 2010b). It is evi-

dent that a geochronological approach is needed for final solution of this problem. Granite Leucosomes in the Gridino Migmatites Attributes of successive metamorphic events are recorded in migmatitic felsic rocks formed under high-pressure conditions.


64

Chapter 2

Figure 2.17. The metagabbro dike on Vorotnaya Luda Island. Undeformed part of the dike has intrusive contacts and crosscuts the felsic gneiss. Extreme deformation of the dike led to segmentation into pods and lenses concordant with the foliation of the host gneisses. Figure is after Mints et al. (2014b); reprinted with permission of Elsevier. Location of figure shown in Figure 2.16.

Granite, enderbite, and quartz veins, which intersect mafic dikes that underwent eclogite metamorphism, determine the upper chronological limit of these metamorphic events. The most important results have been obtained from the study of four key localities: Cape Vargas, the northeastern outskirts of the Gridino Village, Gridin Cape, and Izbnaya Luda Island (Fig. 2.16; Dokukina and Konilov, 2011; Dokukina et al., 2014). Cape Vargas. Cape Vargas is composed of migmatized tonalite gneisses intercalating with amphibole gneisses, which contain numerous amphibolite and eclogite bodies differing in size (from a few centimeters to a few meters) and shape (from equant to flattened) and metagabbro dikes deformed together with the host gneisses (Fig. 2.18). The granite gneisses and metagabbro dikes form a fold package with a gently plunging hinge and inclined axial surfaces. At the southern end of the cape, a segment of the metagabbro dike more than 30 m long reveals crosscutting relations with the host gneiss. In the deformed parts, the dike is amphibolized, while the banding of gneiss and granite is oriented parallel to the dike contact. Deformed mafic bodies and amphibolite boudins apparently represent dike fragments. The dikes of Cape Vargas do not retain primary igneous minerals and

are composed of a garnet-clinopyroxene-plagioclase assemblage with superimposed brown-green amphibole and biotite. Garnet contains rare omphacite inclusions (up to 43 mol% jadeite [Jd]) with a high content of the Ca-Ts end member, which is a relict mineral related to the eclogite stage of metamorphism (Dokukina and Konilov, 2011; Dokukina et al., 2012). Almost all felsic and mafic rocks underwent partial melting during posteclogite decompression. The initial stage of melting is characterized by formation of a phengite-bearing leucosome. Numerous thin veins (from a few to tens of centimeters in thickness) penetrate into gneiss and mafic rocks and contain both restite bodies and unaltered tonalite and mafic fragments (Fig. 2.18A). Petrologic study of the leucosome showed relics related to the earlier high-pressure conditions: Ba-bearing phengite (3.15–3.2 silicon cations per 11 oxygen atoms), K-feldspar and K-Ba–feldspar, myrmekite, and near-solidus symplectic intergrowths of clinozoisite, phengite, and quartz (for more details, see Dokukina and Konilov, 2011). A similar clinozoisite-quartz symplectite was described from diamond-bearing clinozoisite gneiss of the ultrahigh-pressure Kokshetau massif in northern Kazakhstan (Korsakov et al., 2006). The phengite geobarometer (Caddick and Thompson, 2008) yields high-pressure conditions of leucosome crystallization: 15–25 kbar at 650–750 °C. Biotite replacing phengite; grossular garnet and clinopyroxene replacing clinozoisite–quartz symplectite; and breakdown of plagioclase with formation of antiperthite provide evidence for transition from eclogite- to granulite-facies conditions of decompression. The biotite-garnet and garnet-clinopyroxene geothermometers in combination with the garnet-clinopyroxene-plagioclase-quartz geobarometer (Fonarev et al., 1991) indicate high-pressure granulite-facies conditions: 750–800 °C and 10–12 kbar. The last metamorphic event in the leucosome is characterized by replacement of clinopyroxene with amphibole and newly formed biotite and epidote. Northeastern Gridino Village. The northeastern Gridino Village occupies a cape consisting of tonalite gneiss, which contains boudins of retrograde eclogites and is cut through by eclogitized gabbro and gabbronorite dikes (Figs. 2.16 and 2.19A). The structure of the cape and the rocks occurring therein are described in detail by Volodichev et al. (2012). The NW-SE–trending zone of later magmatic recycling and deformation is localized in the northwestern part of the dike. It is ~1–2 m thick and contains an enderbite vein and metasomatic veinlets that have been precisely dated (Table 2.10). A marginal enderbite rim, ~1 cm thick, is enriched in mafic minerals and contains an equilibrium garnet-clinopyroxene-orthopyroxenebiotite-plagioclase-quartz assemblage (Fig. 2.19A, sample 111106) or the same assemblage without orthopyroxene (sample 1111-09). The central part of the enderbite vein is composed of kyanite-garnet rock with sporadic clinopyroxene and orthopyroxene. Garnet from the central part of enderbite vein reveals prograde zoning, with higher Fe and Ca contents in grain cores and with kyanite and omphacite inclusions (up to 21 mol% Jd) in its magnesian rims. Kyanite inclusions were also found in



65

Figure 2.18. (A) Sketch map of Cape Vargas with sampling sites (modified after Dokukina et al., 2012). (B) Undeformed part of the metagabbro dike, from which geochronological sample D17-1 was taken; (C) photo of a deformed and migmatized metagabbro dike; and (D) part of the metagabbro dike penetrated by veinlets of the phengite-bearing leucosome from which geochronological sample D17-4 was taken. White arrows indicate photo locations. Location of figure shown in Figure 2.16.

plagioclase. Pressure and temperature estimates at 750 °C and 12.6 kbar correspond to the high-pressure granulite-facies conditions (Dokukina et al., 2014). Two samples have been collected for geochronological study: one from the root of the vein and another from its continuation in the metagabbronorite body (Figs. 2.19B and 2.19C, sample 1111-06). Metasomatic veinlets confined to the zone of late magmatic recycling and deformation were also found (Figs. 2.19A and 2.19D, sample 1111-08). At a distance from these veinlets, the

dike consists of orthopyroxene eclogite without quartz. In contrast, the veinlets are quartz-bearing symplectic eclogites with linear quartz-biotite-plagioclase or pure quartz streaks. Linear clusters of garnet and rutile develop along quartz streaks. An orthopyroxeneplagioclase corona forms at the boundary between quartz and garnet. Orthopyroxene also occurs in the clinopyroxene-plagioclase symplectite and contacts with K-feldspar. Relics of omphacite (up to 30 mol% Jd) remain in orthopyroxene-clinopyroxeneplagioclase symplectite; kyanite and omphacite inclusions (up to


66

Chapter 2

Figure 2.19. (A) Sketch map northeast of the Gridino Village with sampling sites (modified after Volodichev et al., 2012). (B) Exposure of the root of an enderbite vein from which geochronological sample 1111-06 was collected. (C) Exposure of the continuation of the enderbite vein from which geochronological sample 1111-09 was collected. (D) Exposure of metasomatic veinlets penetrating olivine metagabbronorite dike from which geochronological sample 1111-08 was collected. Opx—orthopyroxene. Figure is slightly modified from Dokukina et al. (2014): Dokukina, K.A., Kaulina, T.V., Konilov, A.N., Mints, M.V., Wan, K.V., Natapov, L.M., Belousova, E.A., Simakin, S.G., and Lepekhina, E.N., 2014, Archaean to Palaeoproterozoic high-grade evolution of the Belomorian eclogite province in the Gridino area, Fennoscandian Shield: Geochronological evidence: Gondwana Research, v. 25, p. 585–613, doi:10.1016/j.gr.2013.02.014. © 2014, reprinted with permission of Elsevier. (Figures 2.20, 2.27, 2.28, and 2.29 are also taken from that article.). Location of figure shown in Figure 2.16.

Fe-Ti eclogite (Fe-Ti metagabbro) (sample S198/713)

Grt-Ky-Qtz vein crossing eclogite (sample S198-1/2)

15

16

Shirokaya Salma 13 Mafic eclogite (metagabbro) (sample S-198/107) 14

12

11

“Granulitic-type” zircons with fir-tree internal zoning

Ovoid multifaceted zircon grains

Almost homogeneous multifaceted rounded “granulitic-type” zircons with an indistinct weak zoning (Figs. 2.24C and 2.24D)

Plagiogranite (garnet [Grt]-kyanite [Ky]quartz [Qtz]) vein crossing eclogite (sample S204/28)

10

2.69

2.695 (upper intercept)

2.70–2.69 (18-point concordant age)

2.72 (discordia age)

1.87 (ten-point concordant age)



1.89 (13 points concordant age)

Garnetite (metamorphosed gabbro with Fe-Ti mineralization) (sample S204/23B)

9

Irregular to tabular grains with inclusion-filled cores similar to zircons from Fe-Ti eclogite, surrounded by broad, transparent, structureless cracked rims: only the rims were dated (Figs. 2.24E and 2.24F) Subhedral grains with clear corerim structures: the cores display fine oscillatory zoning, while rims are featureless: rims and two core populations were dated

2.46 (age peak in histogram) 2.28 (age peak in histogram) 2.1 (age peak in histogram) 1.91 (age peak in histogram)

5 6 7 8

U-Pb, SHRIMP-II, zircon U-Pb, LA-ICP-MS, zircon U-Pb, isotope dilution (ID) thermal ionization mass spectrometry (TIMS), zircon U-Pb, single-grain TIMS

U-Pb, SHRIMP-II, zircon

U-Pb, LA-ICP-MS, zircon

U-Pb, LA-ICP-MS, zircon

Dokukina et al. (2013)

Mints et al. (2010a, 2010b) Konilov et al. (2011)

Kaulina (2010)

Mints et al. (2010a, 2010b); Konilov et al. (2011)

Mints et al. (2010a, 2010b)

TABLE 2.10. AGE DETERMINATIONS ON ROCKS AND METAMORPHIC EVENTS IN THE BELOMORIAN ECLOGITE PROVINCE No. Complex, sample Type of zircon grains Age (Ga) Method References The Salma eclogite association (related to subduction setting) Uzkaya Salma 1 Fe-Ti eclogite Zircons with characteristic patchy 2.89 (one-point concordant age), U-Pb, sensitive Kaulina (2010) (Fe-Ti metagabbro) textures and numerous voids and 2.94 (upper intercept) high-resolution ion (sample S204/2B) mineral inclusions (quartz, plagioclase, microprobe (SHRIMP) clinopyroxene, rutile, calcite, muscovite, II, zircon 2 1.82 (lower intercept) apatite, Al-titanite, allanite, epidote, pyrite, galena) (Figs. 2.24A and 2.24B) 3 2.89 (two-point concordant age) U-Pb, laser ablation– Mints et al. (2010a, inductively coupled 2010b) plasma–mass spectrometry (LA-ICP4 2.82 (age peak in histogram) MS), zircon

Mesoarchean Kola-Karelia continent (Continued)

Metamorphic crystallization under granulite-facies conditions

Magmatic crystallization age Resetting (nonzero Pb loss) of the older zircons Svecofennian mantle plume–related event

Svecofennian mantle plume–related event Svecofennian hightemperature mantle plume– related metamorphic event

Suggested age of magmatic crystallization Svecofennian mantle plume–related event Suggested age of magmatic crystallization Age of the earliest metamorphism Mantle plume– related thermal events

Remarks


67

34

Eclogite-like rock (clinopyroxenite) (sample 21) Island Stolbikha 33 Mafic eclogite (metabasite) (sample 102)

32

Garnetite (sample 48)

Outer rim around core Wide light-gray rim

Dark reworked zone in the cores

Gray reworked zone in core

Dark inclusionfilled core

Oval and rounded zircon crystals, partially with distinct zoning

2.70 (upper intercept) 1.88 (lower intercept)

Pale gray rims

2.72 (15-point concordant age and upper intercept)

2.56–2.38 (three subconcordant estimates) 2.11–2.01 (two subconcordant estimates) 1.93 (lower intercept) 1.90 (five-point concordant age) 1.91 (four-point concordant age)

Dark unzoned cores

Oval and rounded zircon crystals

Rounded transparent zircon grains that usually do not exhibit any zoning

Irregular to shortprismatic grains with inclusion-filled cores similar to zircons from Fe-Ti eclogite, surrounded by broad, transparent, structureless cracked rims

Pale gray rims

2.88 (the weighted mean fivepoint age) 1.88 (the weighted mean six-point age) 2.86 (upper intercept) 2.80 (one-point concordant age) 2.71 (five-point subconcordant age)

Ca. 1.9 (two-points concordant age)

U-Pb, NORDSIM (Nordic ion microprobe facility), zircon U-Pb, SHRIMP-II, zircon




Suggested magmatic nature and an Archean age of the gabbroid protolith of the eclogites Suggested Svecofennian age of eclogite-facies metamorphism (Continued)

Suggested age of the eclogitefacies metamorphism

Volodichev et al. (2004) Skublov et al. (2011a)

Suggested Svecofennian age of the eclogite-facies metamorphism

Suggested Svecofennian age of the eclogite-facies metamorphism

No comments

Suggested age of magmatic protolith rejuvenated owing to Svecofennian metamorphism No comments

Suggested age of magmatic protolith of the garnetite

Suggested magmatic nature and an Archean age of the gabbroid protolith of the eclogites Suggested Svecofennian age of the eclogite-facies metamorphism

Skublov et al. (2011b)

Mel’nik et al. (2013)

Skublov et al. (2011b)

Crystallization under high-pressure granulite- or eclogite-facies conditions Svecofennian tectonothermal event

Svecofennian tectonothermal event Suggested magmatic crystallization

Early Proterozoic metamorphic reworking

Chapter 2

31

30

29

28

27

26

Dark cores

Rounded colorless often cavernous zircon crystals with distinct zoning

25

Mafic eclogite (metagabbro) (sample 46)

Thin rims

Remarks Domains with distinct magmatic zoning show the reliable age of the protolith

(Continued)

68

24

TABLE 2.10. AGE DETERMINATIONS ON ROCKS AND METAMORPHIC EVENTS IN THE BELOMORIAN ECLOGITE PROVINCE No. Complex, sample Type of zircon grains Age (Ga) Method References Chalma (Kuru-Vaara quarry) Ca. 2.9 U-Pb, SHRIMP-II, Shchipansky et al. Cores with Anhedral zircon 17 Mafic eclogite (subconcordant age) zircon (2012b) abundant pores grains with (metagabbro) and inclusions homogeneous or (sample KV-0703) 18 2.82 (four-point concordant poorly patchy-sector age) zoned internal 19 Ca. 2.58 structure (subconcordant ages) 20 Ca. 2.0 (subconcordant ages) 21 Homogeneous Ca. 1.9 (two-point concordant rims age) 2.77 (one-point concordant U-Pb, SHRIMP-II, Shchipansky et al. 22 Mafic eclogite Almost homogeneous Large cores age) zircon (2012b) (metagabbro) multifaceted rounded (sample KV-0706) zircons with unzoned 23 2.72 (four-point concordant structure or weak age) zoning


Eclogite-like garnet–pyroxenite rocks (sample 108)

Oval and rounded zircon crystals with distinct zoning

Phengite-bearing granitic leucosome (sample D17-7)

Eclogitized Femetagabbro dike (sample D17-1)


U–Pb, SHRIMP-II, zircon 2.87 (upper intercept) 2.85 (five points concordant age)

Brownish or colorless, elongate, subhedral zircons with core and rim structure (Figs. 2.26C, 2.26D) Cores of zircon grains with thin oscillatory zoning Colorless rims of zircon grains

2.78 (upper intercept) 2.78 (two points concordant age)

Dokukina et al. (2011b)

U–Pb, LA-IC-PMS, zircon

2.72




Skublov et al. (2011a)

Zircon grains with magmatic type oscillatory zoning

U–Pb, TIMS, zircon

U–Pb, SHRIMP-II, zircon



References

2.64

1.94 (concordant age)

2.71 (discordia age)

2.44 and 2.38 (concordant ages) 2.71 (concordant age) 2.79 (the only discordant age)



3.0 (sub-concordant age)

2.79 (one-point concordant age) 2.74 (upper intercept) 1.90 (lower intercept)

Method

The long-prismatic, brownish zircon grains

Thin colorless rims

The long-prismatic, brownish zircon grains Brownish gray Rare cores with distinct oscillatory short-prismatic and oval-shaped zoning grains (some grains have a “firtree” texture) Main cluster

Zircon crystal fragment with a coarse oscillatory zoning Elongated zircon crystals with magmatic type oscillatory zoning Mainly rounded or ellipsoidal zircon grains with sector, fir-tree, and diffuse oscillatory zoning Two crystal fragments

Pale gray rims

Dark unzoned cores

Age (Ga)

These rims probably resulted from partial melting of the tonalite gneiss at peak conditions (Continued)

Suggested age of mafic dike intrusion

Magmatic crystallization

Svecofennian mantle plume related overprinted amphibolite-facies metamorphism Metamorphic crystallization?


Crystallization induced by the mantle plume–related thermal event Metamorphic crystallization Suggested age of magmatic crystallization


Suggested entrapped grain from the host gneiss Magmatic crystallization age

Suggested magmatic nature and an Archean age of the gabbroid protolith of the eclogites Suggested Svecofennian age of eclogite-facies metamorphism

Remarks


47

Phengite-bearing granitic leucosome (sample D17-4) 45 Granite body crosscutting foliation of the host gneisses Cape Gridin 46 Eclogitized metagabbro dike (sample d44-4)

44

43

42

40 41

39

38

37

36

Cape Vargas

The Gridino eclogitized dike swarm

35

Type of zircon grains

TABLE 2.10. AGE DETERMINATIONS ON ROCKS AND METAMORPHIC EVENTS IN THE BELOMORIAN ECLOGITE PROVINCE (Continued)

Complex, sample

Island Vysoky

No.


69

Metasomatic veinlets crosscutting the eclogitized gabbronorite dike (sample 1111-08)

Enderbite vein (garnetclinopyroxenebiotite-kyanite enderbite, sample 1111-09)

Enderbite vein crosscutting the eclogitized gabbronorite dike (garnetorthopyroxeneclinopyroxenebiotite-kyanite enderbite, sample 1111-06) Predominant group of zircon grains (with inclusions of phengite and omphacite) Pale rims

Short-prismatic crystals with thin oscillatory zoning Long-prismatic crystals with or without zoning Oval-shaped short-prismatic crystals with thin zoning

Some parts of zircon grains, which do not have cavities and inclusions

Zircon grains with an unusual “clumpy” morphology with cavities (Fig. 2.26K)

Rounded or ellipsoidal zircon grains with sector, fir-tree, and diffuse oscillatory zoning (Figs. 2.26I, 2.26J)

thin bright rim

Magmatic subhedral zircons with oscillatory zoning (Figs. 2.26A, 2.26B) Subhedral zircon Magmatic cores grains with distinct (Figs. 2.26C, 2.26D) core-rim structures Rims (Figs. 2.26C, 2.26D)


2.82–2.79 (two subconcordant ages 2.74 (upper intercept)

2.15 (one-point concordant age) 1.87 (one-point concordant age)

1.92 (one-point concordant age) 2.39 (four-point concordant age)

1.98 (one-point concordant age) 2.72 (discordia age)

2.74 ( upper intercept )

2.88–2.83 ( upper intercept )

2.39 (four-point concordant age) 1.91 (five-point concordant age) 3.01 ( upper intercept )






U–Pb, LA-ICPMS, zircon

Method U–Pb, SHRIMP-II, zircon

Slabunov et al. (2008)




Volodichev et al. (2012)

References Dokukina et al. (2014)

Mafic magma crystallization

Slabunov et al. suggested xenogenic nature of these zircon grains

Svecofennian amphibolite-facies metamorphism Percolation of aggressive hot metasomatic fluid through earlier eclogitized mafic dike Hydrothermal (?) alteration linked with regional tectonothermal events

Suggested age of the mafic magma crystallization Svecofennian amphibolite-facies metamorphism The igneous crystallization of the country metatonalite A response of the tonalite gneiss to the thermal effect of the emplaced gabbronorite magma The initial melting of the country metatonalite and the derivation of the granitoid melt under granulitefacies conditions Svecofennian amphibolite-facies metamorphism Partial melting of the country metatonalite under granulite-facies conditions

Ages of grains that could be entrapped from the host rocks (xenogenic zircons)

Remarks Suggested igneous genesis; grain was likely inherited and yield crystallization age of the host tonalite The upper limit for igneous zircon crystallization These rims probably resulted from partial melting of the tonalite gneiss at peak conditions Metamorphic crystallization under granulite-facies conditions

Chapter 2

Suprotivnye Islands 65 Metagabbro that underwent high-pressure 66 amphibolite-facies matamorphism 67

64

63

62

61

60

59

58

57

56

55

Prismatic zircon crystals with oscillatory zoning Unzoned prismatic grains with “clumpy” morphology Short-prismatic and oval-shaped crystals

2.84 (one-point concordant age) 2.77 (upper intercept) 2.70 (upper intercept)

2.72 (age peak in histogram)

2.80–2.78 (age peak in histogram)

2.85 (age peak in histogram)

Age (Ga) Ca. 2.9 (one grain)

70

54

Northeast of Gridino village 52 Eclogitized gabbronorite dike (sample V-16) 53 Short-prismatic zircon crystals with thin oscillatory zoning

Colorless rims of some zircon grains

51

50

Zircon grains with coarse oscillatory zoning (Figs. 2.26E, 2.26F) Colorless rims of zircon grains

Type of zircon grains Colorless subhedral grains with sector or oscillatory zoning (Figs. 2.2A, 2.26B)

TABLE 2.10. AGE DETERMINATIONS ON ROCKS AND METAMORPHIC EVENTS IN THE BELOMORIAN ECLOGITE PROVINCE (Continued)

Complex, sample Granite leucosome (sample d44-1)

49

No. 48



Mesoarchean Kola-Karelia continent 34 mol% Jd) are contained in garnet. It is important to note that symplectite contains Cl-apatite (up to 6.4 wt% Cl), rutile, Ti-rich biotite (up to 7 wt% TiO2), and chains of zircon grains. Veinlets were most likely formed by percolation of brine or melt into eclogitic rock (Dokukina et al., 2012, 2014). The P-T parameters of the symplectic mineral assemblage crystallization correspond

71

to 9–10 kbar at 700–750 °C. Sample 1111-08 was taken for zircon dating from these veinlets (Table 2.10). Cape Gridin. Several large metamorphosed mafic dikes cut Cape Gridin (Figs. 2.16 and 2.20; Dokukina et al., 2014). A thick metagabbronorite dike (Fig. 2.20A) retains relic igneous texture and minerals and displays clearly visible chilled margins with

Figure 2.20. (A) Geological sketch map of Cape Gridin with sampling sites. Ol—olivine. (B) Field photo of intrusive interaction between the eclogitized dikes of later metagabbro and quartz-bearing metagabbronorite. Ky—kyanite; Grt—garnet. (C) Granite leucosome penetrating a metagabbro dike. (D) Field photo of interaction between a granite leucosome vein (sample d44-1), late metagabbro dike (sample d44-4), and olivine metagabbronorite. Figure is slightly modified after Dokukina et al. (2014); reprinted with permission of Elsevier. Location of figure shown in Figure 2.16.


72

Chapter 2

Figure 2.21. (A) Geological sketch map of the central and western domains on Izbnaya Luda Island. (B–D) Photos of the central domain outcrops, Izbnaya Luda Island: B, C—tectonic breccia; D—mafic dikes. Location of figure shown in Figure 2.16.

the host gneiss. This dike is cut by a younger, completely metamorphosed metagabbro dike. Relationships between these dikes are shown in Figures 2.20A and 2.20B. The metagabbro dike has pinch-and-swell morphology; a granite leucosome is visible in a thin zone along its contact. To constrain the age of these dikes, sample d44-4 from the metagabbro dike and sample d44-1 from the granite leucosome vein crosscutting this dike were taken for

geochronological study (Figs. 2.20C and 2.20D). Neither relict igneous textures nor minerals have ever been found in this dike. Metagabbro generally contains an equilibrium high-pressure granulite-facies garnet-clinopyroxene-plagioclase assemblage. The eclogite-stage mineral assemblage is occasionally retained as well. The rock matrix is an aggregate of clinopyroxeneplagioclase symplectite typical of retrogressed eclogites.


Mesoarchean Kola-Karelia continent Omphacite relics with up to 36–42 mol% Jd are also retained within the symplectite and as inclusions in garnet. Omphacite in the matrix is free of inclusions or had oriented quartz needles (rods). Garnet is homogeneous, except for only a slight increase in the iron content at the edges, which can be explained by development of late kelyphitic rims. Sample d44–1 was taken from the granite leucosome vein crosscutting both the metagabbro and metagabbronorite dikes (Fig. 2.20A). The granite leucosome consists of garnet, biotite, plagioclase, K-feldspar, and quartz with minor epidote and scapolite. In one sample, two inclusions of Ti-rich phengite were found in garnet, which is surrounded by a clinopyroxene corona that was formed under high-pressure granulite-facies conditions after emplacement of the vein (Dokukina et al., 2014). Izbnaya Luda Island. Izbnaya Luda Island (Figs. 2.16 and 2.21, inset) is composed of felsic gneiss with bodies of eclogite, symplectite-bearing eclogite and amphibolite, garnetclinopyroxene amphibolite, metaperidotite, metapyroxenite, and microcline granite incorporated therein. The NW- and W-trending foliation is characteristic of the rocks occurring in the central domain (Figs. 2.21A and 2.22). This banded gneiss consists of layers enriched in biotite and light layers rich in quartz and feldspar. Gneiss foliation is cut by fractures oriented in different directions and with slip varying from several centimeters to a few meters. Movements along these fractures resulted in formation of tectonic breccia (Figs. 2.22B and 2.22C). Gneiss contains fragments of garnet (±clinopyroxene) amphibolite, eclogite, tonalite, and pink K-rich granite 10 cm to 5 m in size. The elongated mafic inclusions are crosscut by migmatite veins. The granite and tonalite inclusions are angular or stretched along the foliation. Fracture networks are irregularly distributed in space. Relics of early folds are retained within these networks. Tectonic breccia is crosscut by thin (a few millimeters to a few centimeters) metapseudotachylyte veinlets, which are represented by fine-grained aplite-like material (Fig. 2.22B). Relatively large-scale metapseudotachylyte veins contain small fragments of gneiss and K-rich granite. The networks of thin metapseudotachylyte veinlets surround gneiss and granite fragments (Fig. 2.22C; Dokukina and Konilov, 2011; Dokukina, 2015). Metapseudotachylyte is the garnet–biotite–plagioclase– K-feldspar–quartz ± clinopyroxene ± hornblende rock, the bulk composition of which corresponds to the average composition of tectonic breccia. The Sm/Nd model age tDM varies from 2787 to 3148 Ma (Dokukina et al., 2011a, 2011b). Thus, metapseudotachylyte veins are important evidence for a high-velocity seismogenic deformation of the Mesoarchean continental crust. Mafic dikes of the Gridino swarm crosscut the tectonic breccia together with metapseudotachylyte veinlets (Fig. 2.21). 2.4.2. Pressure-Temperature Variations Thermobarometrical study of the Belomorian eclogites was performed using a multicalibration approach based on the

73

consistent system of geothermometers and geobarometers of the TPF program (Appendix II-3 [see end of volume]) developed at the Institute of Experimental Mineralogy, Russian Academy of Sciences (Fonarev et al., 1991, 1994; Konilov, 1999; Maaskant, 2004). The detailed descriptions of the results of thermobarometric investigations are available from Volodichev et al. (2004), Mints et al. (2010b), Konilov et al. (2011), Dokukina and Konilov (2011), Shchipansky et al. (2012a), and Dokukina et al. (2014). Salma Eclogite Association The main features of the metamorphic evolution of the Salma eclogites are shown in Figure 2.23. Prograde evolution. Diaspore inclusions in spinel from the Uzkaya Salma piclogite suggest that the protolith, before being subducted, was above sea level and exposed to lateritic weathering (for more details, see Konilov et al., 2011). A recent analog of such a rock has been described in the metabauxites from Samos and Naxos Islands, Greece (Feenstra, 1997; Urai and Feenstra, 2001). An ancient laterite containing oxides or oxyhydroxides was previously known in the Hekpoort paleosol of the Transvaal Supergroup (ca. 2.2 Ga) in South Africa (Beukes et al., 2002). Our data imply that an oxygenated atmosphere existed ca. 2.8 Ga, though it is not widely accepted. Additional information about the earliest stages of metamorphism is provided by mineral inclusions retained in the prograde-zoned atoll garnets from the Uzkaya Salma locality. The low-pressure and lowtemperature minerals are pumpellyite, actinolite, albite, zoisite, and chlorite, which occur either as isolated mineral phases or as relict mineral assemblages armored by the atoll garnet. The textural evidence clearly suggests that the pumpellyite-actinolite assemblage was replaced later by hornblende. Moreover, there is textural and compositional evidence indicating that these inclusions were entrapped during prograde growth of the atoll-type garnet (Konilov et al., 2011). Thus, these phases can be interpreted as a manifestation of the prograde mineralogy (Page et al., 2003, 2007), and they indicate that the mafic protolith has passed through the pumpellyite-actinolite metamorphic facies during its prograde path. The low-temperature alteration in the gabbroic suite of the slow-spreading Southwest Indian Ridge is typically confined to fractured regions, where intense alteration of the host rocks is observed adjacent to veins and veinlets filled with smectite, smectite-chlorite mixed layer minerals, or chlorite ± calcite ± zeolite ± sulfide ± Fe-oxyhydroxide (Dick et al., 2000; Bach et al., 2001). Most of the minerals listed here are commonly considered to be indicators of seafloor metamorphism (Nozaka et al., 2008). Several reports on similar low P-T mineral assemblages incorporated into garnet from Phanerozoic eclogites have been recently published (Enami et al., 2004; Yang and Powell, 2008). A later prograde event has been documented by the clinopyroxene that armors chlorite in the cores of garnets from the Salma eclogites. The maximum temperature calculated from the garnet-clinopyroxene assemblage was 636 °C at 10 kbar reference pressure. These aspects of the Salma eclogites


74

Chapter 2

Figure 2.22. Geological sketch map (A) and outcrop photos (B, C) of the typical pattern of the tectonic breccia and metapseudotachylyte veinlets in the central domain, Izbnaya Luda Island.

have been recently described in detail by Mints et al. (2010b), Konilov et al. (2011), and Shchipansky et al. (2012a). Eclogite event. Although the studied localities contain mainly retrogressed eclogites with almost complete replacement of the primary matrix omphacite by diopside-plagioclase symplectite, petrographic study suggests that in all cases the mineral assemblages of high-pressure stages were plagioclase-free. The minimum pressure for the eclogite stage was estimated as a func-

tion of the Jd content in clinopyroxene coexisting with quartz and albite. Inclusions of omphacite were observed within garnet in a number of thin sections, allowing us to constrain the peak P-T conditions of this assemblage. For the Shirokaya Salma eclogite, the peak pressure is inferred to be ~13 kbar at a temperature of 700 °C. The samples from the Uzkaya Salma eclogite yielded pressures of 13–14 kbar at a temperature range 700–750 °C. The samples from the Chalma (Kuru-Vaara quarry) eclogites give



Figure 2.23. The pressure-temperature-time (P-T-t) evolution of the Salma and Gridino eclogite associations and comparison of the apparent P-T conditions along Phanerozoic subduction zones, modified after Mints et al. (2014b). Reprinted with permission of Elsevier. PP—prehnite-pumpellyite, EBS—epidote blueschist, LBS—lawsonite blueschist, GS—greenschist, EA—epidote amphibolite, A—amphibolite, GA—garnet amphibolite, PG—pyroxene granulite, and GG—garnet granulite. P-T-t paths of the Salma and Gridino eclogite associations are summarized from Kaulina et al. (2007, 2010), Dokukina et al. (2009, 2012, 2014), Mints et al. (2010a, 2010b), Kaulina (2010), and Dokukina and Konilov (2011). Calculated P-T paths and metamorphic conditions encountered by oceanic crust subducted beneath Cascadia, southern Mexico, southwest Japan, and northeast Japan are after Peacock et al. (2002). Metamorphic facies are after Peacock et al. (1994). Chl—chlorite, Amp—amphibole, Zo—zoisite, Ky—kyanite, Sil—sillimanite, And—andalusite.

75


76

Chapter 2

slightly higher temperatures of 750–775 °C and pressures of 14.0–14.3 kbar (Mints et al., 2010b; Konilov et al., 2011; Shchipansky et al., 2012a). All these pressures should be regarded as minimums due to disequilibrium of plagioclase with the aforementioned mineral assemblage. Relict kyanite occurs among the diopside-plagioclase symplectites, indicating that omphacite was abundant in the peak-pressure assemblage, although the composition of the peak-pressure omphacite formerly associated with kyanite remains unknown. Omphacite with >32 mol% Jd has been found as a single inclusion within garnet. The relict matrix omphacite (up to 28 mol% Jd) retained in a sample from the Kuru-Vaara quarry gives evidence for quartz exsolution (Konilov et al., 2011; Shchipansky et al., 2012a). Similar exsolution microstructures have been found in eclogite from several ultrahigh-pressure metamorphic belts, including the Kokshetau Massif in Kazakhstan (Katayama and Maruyama, 2009), Pohorje in the eastern Alps, Slovenia (Janák et al., 2004), Alpe Arami, Switzerland (Dobrzhinetskaya et al., 2002), and the Blumenau eclogite, Erzgebirge, Germany–Czech Republic (Chopin and Ferraris, 2003). Most authors consider these features to be diagnostic of ultrahigh-pressure metamorphism. Retrograde evolution. The eclogite decompression has promoted omphacite breakdown and formation of diopside-plagioclase symplectite. A pressure range from 10 to nearly 13 kbar at 750 °C was inferred for symplectitic mafic and Fe-Ti eclogites from the Shirokaya Salma and Uzkaya Salma localities. Lower values of these parameters were obtained for garnet rims and clinopyroxene-plagioclase-quartz coronas (7–10 kbar and 670–700 °C). The temperature of hornblende- and biotitebearing assemblages is generally estimated at 650 °C to 700 °C at inferred pressures of 7–9 kbar. It remains unclear whether these assemblages represent continual cooling with a slight decrease in pressure or a later metamorphic event. Thus, the retrograde paths of the eclogite bodies generally pass via garnet-amphibolite to upper amphibolite facies, thus representing almost isothermal decompression. Another evolutionary path has been recognized for eclogites from the Shirokaya Salma and Chalma (Kuru-Vaara quarry) localities, which experienced decompression coupled with heating. The presence of orthopyroxene intergrown with the plagioclase-diopside symplectite implies retrogression via granulite facies at a temperature of 730–780 °C. An increase in temperature up to medium-grade granulite facies after attainment of peak pressure has been reported from several Phanerozoic subduction-related eclogites (Page et al., 2003, and references therein) and from almost all known Precambrian occurrences (Möller et al., 1995; Collins et al., 2004; Carlson et al., 2007; Mosher et al., 2008). Gridino Swarm of Eclogitized Dikes The main features of the metamorphic evolution of the Gridino eclogitized dikes are shown in Figure 2.23. Petrographic examination of thin sections clearly shows that all studied dikes underwent successive metamorphism from eclogite via highpressure granulite to amphibolite facies. They also display a

structural heterogeneity, which is more pronounced in the larger bodies. In the undeformed thin dikes, the granulite garnetorthopyroxene-clinopyroxene (not omphacite) assemblage, with rare relics of clinopyroxene-plagioclase symplectite, is predominant. In the larger dikes, 25–30 m thick, all stages of the metamorphic evolution are identified (Dokukina and Konilov, 2011; Dokukina et al., 2012, 2014). Magmatic stage. Igneous structures and mineral assemblages are commonly retained in the central parts of the dikes, typically containing orthopyroxene and pyroxenes with inverted pigeonite and augite exsolution textures. A crystallization temperature of ~1030 °C has been estimated for quartz-bearing dikes and 1200 °C for olivine-bearing dikes. A minimum temperature of 600 °C at 5 kbar for breakdown of augite and pigeonite corresponds to a postmagmatic stage of isobaric cooling. Our results are consistent with the P-T estimates reported by Egorova and Stepanova (2012) for gabbronorite intrusions: 3.5–5.5 kbar and 1050–1200 °C. Incipient eclogitization. A garnet corona with inclusions of kyanite, plagioclase, clinopyroxene, and quartz formed at the contacts between magmatic pyroxene and plagioclase corresponds to the earliest stage of eclogitization. Eclogite event. As the eclogite assemblage evolved further, coronitic garnet with kyanite inclusions completely replaced igneous plagioclase. The rock matrix is an aggregate of orthopyroxene-clinopyroxene-plagioclase symplectite, which is typical of retrogressed eclogites. In rare cases, omphacite relics (with up to 36–42 mol% Jd) are retained in symplectite and as inclusions in garnet together with kyanite and quartz. Some omphacite grains in the matrix contain oriented quartz needles, which are usually interpreted as an exsolution texture of highsilicic clinopyroxene during decompression of ultrahigh-pressure eclogite (Katayama et al., 2000; Tsai and Liou, 2000). The apparent ultrahigh-pressure formation conditions of Gridino rocks were discussed in much detail by Dokukina and Konilov (2011) and Perchuk and Morgunova (2014). The minimum pressure of 16.0–17.5 kbar was obtained using the jadeite geobarometer (Holland, 1980). The quartz-free rutile-garnet-orthopyroxeneomphacite assemblage (up to 24 mol% Jd) in the metagabbronorite was in equilibrium at a pressure of 22 kbar according to the garnet-orthopyroxene geobarometer (Harley, 1984); see Dokukina et al. (2014) for more details. Granulite event. Orthopyroxene occurs in the clinopyroxeneplagioclase symplectite as evidence for a retrograde metamorphic path in granulite-facies P-T space (Page et al., 2003; Groppo et al., 2007); orthopyroxene also forms coronas around garnet; and it is found within granoblastic orthopyroxene-clinopyroxene domains. In some places, dikes have been totally recrystallized under granulite-facies conditions with formation of garnet–twopyroxene paragenesis in metagabbro. Scrutiny of thin sections of mafic rocks allowed us to suggest that there were at least two granulite-facies events in the history of the Belomorian eclogite province: (1) a posteclogite decompression stage at 10–13 kbar and 750–800 °C, and (2) subsequent overprint of heating (Dokukina


Mesoarchean Kola-Karelia continent et al., 2014). The second stage of the granulite-facies reworking can be inferred from a metagabbronorite dike in the northwestern Gridino Village: The prograde growth of orthopyroxene around actinolitic amphibole is evidence for a temperature increase from 550–600 °C to 750 °C under nearly isobaric (~10 kbar) conditions. The P-T conditions of the origin of the metasomatic veinlets also correspond to 700–750 °C at 9–10 kbar. Amphibolite-facies recrystallization. Dike contacts with host gneisses are usually emphasized by development of an amphibolite assemblage in a zone up to 10 cm wide. Amphibolite-facies assemblages also occur along late fractures, felsic pegmatites, and carbonate veins. Temperature and pressure estimates for amphibolite-grade metamorphism of mafic and felsic rocks are 530–660 °C and 7.9–9.6 kbar (Sibelev et al., 2004; Volodichev et al., 2004; Dokukina and Konilov, 2011). Thermal impact ca. 2.4 Ga. Zircon ca. 2.4 Ga in age from veinlets cutting the eclogitized metagabbronorite dike on the cape at the northeastern end of Gridino Village is of metasomatic origin and probably crystallized from the hot chlorine-bearing brine-melt, which percolated through the dike at a high temperature (Dokukina et al., 2012, 2014). Obvious petrologic evidence for an increase in temperature is the prograde growth of orthopyroxene around low-temperature amphiboles in an amphibolite band of the olivine gabbronorite dike. The possibility of reheating in polymetamorphic high-grade complexes was discussed in the literature (e.g., Mints and Konilov, 1998; Fonarev and Konilov, 2005; Goncalves et al., 2004; Perchuk, 2005). The retrograde zigzag pulsatory P-T path with a staggered declination in the direction of higher temperatures is dated at ca. 2.4 Ga (Fig. 2.23); see Figure 19 in Dokukina et al. (2014) for more details. This event of plume nature operated widely throughout the Belomorian orogen. A superplume was formed ca. 2.5–2.4 Ga in the continental mantle. The superplume predetermined the origin of a large igneous province (LIP) in the Kola-Karelia region, the partial melting of the crustal rocks, and the derivation of anatectic granite and granulite-facies metamorphism (Mints et al., 2010b, and references therein). Gridino Host Gneisses The mineralogy of the host gneisses is generally uniform: biotite, garnet-biotite, or garnet-hornblende tonalite plagiogneisses. Typically, the gneisses are migmatized, and eclogitefacies assemblages have not been found, which is characteristic of the Belomorian eclogite province and other eclogite-bearing terranes. The quartz-rich rocks are more susceptible to dynamic recrystallization and retrograde metamorphism than rocks of mafic composition (Koons and Thompson, 1985). In rare cases, granulite-facies biotite-garnet or kyanite-garnet mineral assemblages are noted in the gneisses. In general, the temperatures and pressures are consistent with those of the eclogites retrogressed under amphibolite-facies conditions. Granulite-facies mineral assemblages in the gneisses correspond to temperature and pressure of 750–800 °C and 10–12 kbar (Dokukina and Konilov, 2011).

77

2.4.3. Geochronology The Belomorian eclogite province provides unique evidence for a Mesoarchean and Neoarchean age of subduction and collision events. However, though the geochronological data obtained by different authors generally coincide, there are appreciable discrepancies in interpretation of the results (Table 2.10). Three different interpretations concerned with the nature and age of the metamorphic events in the Belomorian eclogite province are currently discussed, particularly in Russian journals. The first interpretation suggests three distinct episodes of metamorphism under eclogite-facies conditions: two Archean events related to subduction processes 2.82–2.80 Ga and 2.72– 2.70 Ga, followed by a distinct eclogite-facies event that caused metamorphism of the Paleoproterozoic Gridino mafic dike swarm ca. 2.45–2.40 Ga (Volodichev et al., 2004; Slabunov et al., 2011b; Shchipansky et al., 2012b). The second interpretation assumes one eclogite-facies metamorphic event within the Belomorian province ca. 1.9 Ga (Skublov et al., 2010a, 2010b, 2011a, 2011b). In the third interpretation developed by the authors of this book (Dokukina and Konilov, 2011; Dokukina et al., 2010, 2012, 2014; Mints et al., 2010b, 2014b), the Belomorian eclogite province is considered to be a combination of the subduction-related eclogites and the eclogitized mafic dikes. Both are interpreted as products of closely linked events within a single tectonomagmatic cycle. The magmatic processes and the major eclogite-facies events were constrained to a short time interval between ca. 2.9 and ca. 2.8 Ga. A detailed justification for the Archean age of the Gridino mafic dikes swarm, age of eclogite-facies metamorphism, and subsequent metamorphic events was given by Dokukina et al. (2012, 2014). The geochronological data presented here have already been published elsewhere. The U-Pb dating and Lu-Hf isotopic analyses of zircons from various types of the Salma eclogites and rocks from the Gridino area have been carried out at the GEMOC Key Center, Macquarie University (Sydney, Australia) using the laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) technique (Mints et al., 2010a, 2010b; Dokukina et al., 2012, 2014). The zircons from most other samples, including eclogites of the Salma and Gridino associations, host gneisses, and crosscutting felsic veins, were studied using the SHRIMP II system at VSEGEI (St. Petersburg, Russia; Kaulina et al., 2007; Kaulina, 2010; Slabunov et al., 2008, 2011b; Dokukina et al., 2009, 2010, 2012, 2014; Dokukina and Konilov, 2011; Skublov et al., 2010a, 2010b, 2011a, 2011b; Konilov et al., 2011; Shchipansky et al., 2012b). Some zircons were also dated by NORDSIM at the National Museum of the Natural History (Stockholm, Sweden; Volodichev et al., 2004) and at the Geological Institute of the Kola Science Center (Apatity, Russia; Kaulina et al., 2007; Dokukina and Konilov, 2011; Dokukina et al., 2012, 2014). Herwartz et al. (2012) attempted to estimate the age of eclogite-facies metamorphism using the Lu-Hf method applied to garnet. Measurements were carried out on a Finnigan Neptune multicollector (MC) ICP-MS at the


78

Chapter 2

Steinmann Institute in Bonn, Germany. In this section, we integrate the available geochronological data. Salma Eclogite Association Recent investigations make it possible to distinguish several zircon populations related to the successive events during the evolution of the Salma eclogites (Fig. 2.24). (1) Zircons with patchy textures and numerous voids and mineral inclusions (quartz, plagioclase, clinopyroxene, rutile, calcite, muscovite, apatite, Al-titanite, allanite, epidote, pyrite, galena; Figs. 2.24A and 2.24B) and magmatic-type high concentrations of trace elements, including HREEs. The most reliably estimated crystallization age is 2.90–2.89 Ga. A concordant data point at 2.82 Ga probably represents the age of peak metamorphism, although there is as yet no evidence for ruling out this estimate as characterizing the earliest or retrograde metamorphism. The initial 176Hf/177Hf isotope ratios in zircons show negligible variations when compared to the range of measured U-Pb ages. This implies that the U-Pb systematics were disturbed, whereas the Hf isotope system remained robust (Fig. 2.25). The Nd model tDM ages fall within the 2.88–2.66 Ga time interval, indicating that transfer of the melt into the crust immediately followed its separation from the mantle source. The εHf values vary widely from –16 to +8 (Mints et al., 2010a). This variation mainly reflects the different U-Pb ages used for εHf calculation, where zircons that experienced significant Pb loss or complete resetting of the U-Pb system show more negative εHf values. The eclogite-facies metamorphism led to a gradual decrease in the abundance of trace elements and to a reduction of the Th/U ratio from 2.5–3.0 to 0.1–0.2. The crystal rims and recrystallized zones within the zircon crystals are related to the late events dated at 2.5–2.4, 2.3–2.2, and 2.0–1.9 Ga (Mints et al., 2010b). Zircons of this type were first extracted from Fe-Ti eclogites

Figure 2.24. Specific zircon populations that are related to the successive events in the evolution of the Salma eclogite association, after Mints et al. (2014b). Reprinted with permission of Elsevier. The length of the zircon crystals is in the 0.15–0.25 mm range. Refer to explanation in the text.

investigated at the Uzkaya Salma locality. Subsequently, morphologically similar zircons of the same age have been recognized in the cores of crystals in the “normal” mafic eclogites from the Chalma (Kuru-Vaara quarry) locality (Shchipansky et al., 2012b). The aforementioned morphology, structure, and geochemistry of zircons bear evidence for magmatic crystallization involving highly fractionated residual melt or fluid in deficient

Figure 2.25. Plot of U/Pb age vs. 176 Hf/177Hf, in which Hf shows relatively narrow horizontal distribution of points that suggests resetting of U-Pb system, where zircons with younger U-Pb ages have clearly been affected by nonzero Pb loss (ca. 1.9 Ga). TDM—model age. Figure is modified after Mints et al. (2010a). DM—depleted mantle, CHUR—chondritic uniform reservoir.


Mesoarchean Kola-Karelia continent space. As conditions changed, the zircon crystals were partially recrystallized, although the zircon-bearing rock apparently did not undergo partial melting or alteration induced by fluid activity. Similar features of recrystallized late-magmatic zircons in metamorphic rocks formed at high pressure and/or temperature have been noted elsewhere (Hoskin and Black, 2000; Tomashek et al., 2003; Kaczmarek et al., 2008, and references therein). (2) Almost homogeneous multifaceted rounded granulitic zircons devoid of zonal internal structure or with vague zoning (Figs. 2.24C and 2.24D). Low concentrations of trace elements (especially Y and HREEs) in these zircons suggest crystallization in equilibrium with garnet. Two types of granulitic zircons have been recognized: (i) colorless, low-U zircon with Th/U ranging from 0.16 to 0.47, and (ii) dark-brown, high-U (up to 900 ppm) grains with lower Th/U ranging from 0.12 to 0.33. Based on their morphology, these zircons appear to be metamorphic and could have crystallized under high-pressure granulite- or eclogite-facies conditions (Corfu et al., 2003; Bibikova et al., 2004). The close association of both zircon types is consistent with their crystallization under granulite-facies conditions (Kaulina, 2010). The LA-ICP-MS and SHRIMP ages of these zircons are coincident at 2.72–2.70 Ga. In the 176Hf/177Hf versus U-Pb age diagram (Fig. 2.25), granulitic zircons are generally located between the CHUR and depleted mantle (DM) evolution lines (Mints et al., 2010a). The εHf values vary from +1 to +4, while the Nd model (tDM) age is 2.77–2.90 Ga. The significant difference (up to 180 m.y.) between the Lu-Hf model age and the U-Pb age of magmatic crystallization provides evidence that granulitic zircon crystallized a long time after formation of the igneous host rocks. The lower intersection of the concordia with discordia yields 1.91 Ga as the age of the final metamorphic event. Zircons of this type dated at 2.72–2.70 Ga were found in the Shirokaya Salma (Mints et al., 2010a; Kaulina et al., 2007, 2010; Kaulina, 2010), Chalma (Kuru-Vaara quarry) (Shchipansky et al., 2012b), and Gridino (Volodichev et al., 2004) eclogites. All of them were reworked under granulite-facies conditions, including eclogitized mafic dikes (Dokukina and Konilov, 2011; Dokukina et al., 2012, 2014). The corresponding varieties of rocks can be denoted as eclogite-granulites. It is noteworthy that the zircons of populations (1) and (2) have not been found together in the same samples. We suppose that granulitic zircons were crystallized utilizing Zr released from the eclogite-forming minerals in rocks, which originally had no magmatic zircons. (3) Zircons from the garnetite are irregular to tabular. Many crystals have cores with numerous inclusions similar to the zircons of population (1) collected from the Fe-Ti eclogite surrounded by wide, transparent, and structureless cracked rims (Figs. 2.24E and 2.24F). The cores of zircon crystals in garnetite from the Chalma site (Kuru-Vaara quarry) were formed 2.86 Ga (Mel’nik et al., 2013). Some crystals have been completely altered. The geochemical features and isotopic systematics of the zircon crystals, including their older relict cores, were substantially modified during Paleoproterozoic metamorphism ca. 1.89 Ga (Mints et al., 2010a, 2010b) or 1.93–1.92 Ga (Mel’nik

79

et al., 2013). Both estimates coincide within the limits of analytical uncertainty. Unlike the other samples, this population shows a wide εHf range of −14 to +14 (Fig. 2.25). The lower values are similar to those of the Fe-Ti eclogite, whereas the higher values are well above the depleted mantle evolution line (Mints et al., 2010a). The wide range of εHf values suggests that zircons in garnetite were formed owing to the local breakdown of older Zrbearing minerals, as has been established for zircons from garnetite xenoliths in basalts of the Four Corners area in the United States (Smith and Griffin, 2005). Drastic modification of the isotopic and geochemical properties of zircons may be related to the formation of zones of fluid permeability in the coarse-grained garnetites. They might play the role of distinctive ball bearings within the weakened zones, where the Paleoproterozoic shear deformation was localized. Similar attributes of zircon recrystallization in eclogites have been described in two samples from the Chalma locality (Kuru-Vaara quarry), which contained multifaceted round and oval-shaped zircons grains with cores enriched in inclusions and dated at 2.88 Ga (Skublov et al., 2011b). The outer rims of zircon grains from eclogite and the zircons of similar morphology from eclogite-like garnet-amphibole-clinopyroxene rock are 1.91 and 1.84 Ga in age, respectively. It is noteworthy that the recrystallization event under granulite-facies conditions in the samples described by Skublov et al. (2011b) is based only on the roundto oval-shaped morphology of the zircons resembling the crystals of population (1) rather than on the isotopic geochronological characteristics of zircon. As was mentioned already, Shchipansky and his coauthors also dated granulitic zircons of population (2) in the samples from the same locality. These zircons and morphologically similar zircons from other localities in the Belomorian province yielded exactly the same age (ca. 2.72 Ga). As in the case of zircons from garnetites, the isotopic geochronological characteristics of the zircons in the samples analyzed by Skublov and his coauthors apparently have been completely reworked during the latest metamorphic event. Gridino Eclogitized Dike Swarm Zircons from the mafic eclogitized dikes of the Gridino swarm show a wider morphological diversity when compared with those of the Salma eclogites. This makes it more difficult to correlate individual zircon populations with specific magmatic and/or metamorphic events and engenders considerable discussion on the provenance of different zircon populations in the eclogitized dikes, as well as on age of the dikes and their sequential reworking under the eclogite-, granulite-, and amphibolitefacies conditions. To date, the results of detailed geochronological studies of three eclogitized dikes have been published: at Cape Vargas (Dokukina and Konilov, 2011; Dokukina et al., 2009, 2012, 2014; Fig. 2.18), from the cape in the northeastern outskirts of Gridino Village (Volodichev et al., 2009; Slabunov et al., 2011b; Fig. 2.19), and in the shore outcrop to the southeast of Gridino Village (Gridin Cape; Dokukina et al., 2014; Fig. 2.20). In


80

Chapter 2

addition, a gabbro intrusion located on the Suprotivnye Islands in the immediate vicinity of the Gridino dike swarm has also been studied and dated (Slabunov et al., 2008), as well as several felsic veins crosscutting the eclogitized dikes (Dokukina et al., 2009, 2010, 2012, 2014). Whether the eclogite-facies metamorphism in the Gridino dike swarm is of Paleoproterozoic or Neoarchean age is the subject matter of recently holding debates. Revelation of early Paleoproterozoic zircons in the metasomatic veinlets (Dokukina et al., 2012, 2014) intersecting the olivine metagabbronorite dike previously described by Volodichev et al. (2009) and Slabunov et al. (2011b) is of particular interest in this regard. The data on zircon populations from the mafic eclogitized dikes and the crosscutting felsic veins (for more details, see Dokukina et al., 2014) allow us to delineate the following sequence of events (Fig. 2.26). (1) Subhedral oscillatory zoned magmatic zircons (Fig. 2.26) sporadically recognized in both mafic dikes and crosscutting felsic veins yielded 206Pb/207Pb ages of 3.0–2.91 Ga. These ages are older than a maximum age suggested for the protoliths of the Salma eclogites and most probably correspond to the age of embryos, or the earliest continental fragments found in the Fennoscandian Shield (Mints et al., 2010b; Dokukina et al., 2014). (2) Stubby prismatic to bipyramidal and elongated subhedral zircons dated at 2.87–2.82 Ga (Fig. 2.26) have been found in some mafic dikes and also in the Archean gabbro intrusions on the Suprotivnye Islands (Slabunov et al., 2008). These zircons differ in their morphology, internal structure, and age from the zircons in the country gneisses and granite gneisses (Belousova and Natapov, 2011, personal commun.; see also Mints et al., 2010b). Therefore, they hardly could have been entrapped from granite gneisses by mafic melt and, in contrast, crystallized from this melt itself. Thus, these zircons, most probably, date magmatic crystallization of gabbro intrusions in Suprotivnye Islands and the Gridino dike swarm. (3) Elongated oval-shaped zircons, mainly with large dark cores and light- or gray-colored rims in cathodoluminescence (CL) images, dated at 2.86–2.83 Ga (Figs. 2.26E and 2.26F) and 2.82–2.78 Ga (Figs. 2.26G and 2.26H) occur in both eclogitized mafic dikes and in crosscutting migmatite felsic veins on the Gridin Cape. Geological relationships between the dikes and veins together with morphological and compositional similarity of the zircons are consistent with close interaction among the intrusion of mafic magma, generation of felsic melts, and their mutual crystallization and recrystallization under high-pressure (up to eclogite-facies) conditions. The existence of two zircon populations of the same type indicates successive repetition of the crystallization-recrystallization process. We conclude further that younger zircons correspond to the eclogite-facies event in the magmatic and metamorphic evolution of the Gridino dikes. Ages within the time interval of 2.9–2.8 Ga have also been obtained for cores in the zircon crystals, which belong to the younger populations on the basis of characteristic morphology, geochemistry, and geochronological data (see following).

Figure 2.26. Specific zircon populations that are related to the successive events in the evolution of the Gridino eclogite association, after Mints et al. (2014b), based on data from Dokukina et al. (2014). Figure reprinted with permission of Elsevier. The length of the zircon crystals is in the 0.2–0.25 mm range. Mineral abbreviations in K: Qtz—quartz, Opx—orthopyroxene, Pl—plagioclase, Bt—biotite, Cl-Ap—chloro-apatite. Refer to explanation in the text.


Mesoarchean Kola-Karelia continent (4) The metagabbroic dike at Cape Vargas contains characteristic round- and oval-shaped zircons similar to the granulitic zircons from the Salma eclogites. In both cases, these zircons yielded the same age of 2.72–2.70 Ga. Zircon crystals of similar age (2.74–2.72 Ga) were revealed in the enderbite vein crosscutting the metagabbronorite dike 1111 (in the northeastern outskirts of Gridino Village). The main zircon population in this sample consists of rounded or ellipsoidal colorless or pale brownish grains, 100 × 100 or 100 × 200 μm in size, with soccer-ball morphology and sectorial and fir-tree zoning in CL (Fig. 2.26I), which are typical of granulite-facies zircon in the presence of partial melt (Vavra et al., 1996; Rubatto, 2002; Corfu et al., 2003; Whitehouse and Kamber, 2003). This means that magmatic and metamorphic events were coeval, which is typical of granulite complexes, where the formation of synmetamorphic granitoids is a common situation. An important observation has been made in sample 1111-09 of garnet-clinopyroxene-biotitekyanite enderbite (Dokukina et al., 2014): One of zircon grains from the enderbite vein contains polycrystalline inclusions of omphacite-tschermakite solid solution with phengite, biotite, and quartz (Fig. 2.27). These relationships mark the upper age limit of the eclogite-facies metamorphism (Kröner et al., 2006). Finally, the phengite-bearing leucosome cutting the metagabbro dike at Cape Vargas contains subhedral zircon crystals with a distinctly recognized core and rim. The cores display relict oscillatory zoning, and some crystals have the fir-tree cores (Fig. 2.26J) typical of zircons from high-grade metamorphic

81

rocks (Corfu et al., 2003). A single cluster of nearly concordant estimates yielded an upper-intercept age at 2.71 Ga (Dokukina et al., 2012). (5) Zircons recovered from the thin metasomatic veinlets that penetrate the metagabbronorite dike in the northeastern outskirts of the Gridino Village (early Paleoproterozoic dike B-16 according to Volodichev et al., 2009; Slabunov et al., 2011b) belong to the population characterized by a concordant U-Pb age of 2.39 Ga. These zircons have a distinctive lumpy morphology with voids as relics of fluid inclusions and without internal zoning as evident from CL images (Fig. 2.26K). These zircons contain very high concentrations (wt%) of Th (0.2–1.7), U (0.5–0.9), Y (0.5–3.0), and Hf (0.7–0.9) and also contain numerous inclusions of orthopyroxene, clinopyroxene, Ti-rich biotite, rutile, quartz, and Cl-apatite, which are characteristic of symplectite assemblages and crystallized from fluid (Dokukina et al., 2012, 2014). It is noteworthy that similar zircons from the same dike have previously been interpreted as typical magmatic zircons that date mafic intrusions (Volodichev et al., 2009; Slabunov et al., 2011b). We believe that our results allow us to elaborate upon this understanding; accordingly, we propose that these zircons, hosted in a high-temperature zone, were formed after injection of the dikes and their transformation under eclogite- and late granulite-facies conditions. (6) In the rocks reworked by late processes, which obliterated the eclogite- and granulite-facies mineral assemblages and where they occur only as relics, the Archean magmatic and metamorphic zircon cores are usually surrounded by the colorless low-U rims dated at ca. 1.9 Ga. Dating of single zircon grains from such rocks throughout the Belomorian belt tends to yield the same age (Bibikova et al., 2004; Slabunov et al., 2006a; Mints et al., 2010b). (7) All zircon grains from the granite leucosome (sample d44–1, Cape Gridin) have similar Hf isotope compositions (Fig. 2.28) and show a relatively narrow array of points with εHf close to the CHUR evolution line. Different 207Pb/206Pb ages at similar Hf isotope ratios suggest that the U-Pb system was reset without disturbance of the Hf isotopic composition. 2.4.4. Discussion

Figure 2.27. Backscattered electron image of polycrystalline “nanoenderbite” inclusion with phengite and omphacite in 2.72 Ga zircon from the enderbite vein, after Dokukina et al. (2014). Reprinted with permission of Elsevier. Mineral abbreviations: Qtz—quartz, Ph— phengite, Bt—biotite, Omp—omphacite.

Origin of Protoliths Interpretation of geochemical data allows us to infer an intraoceanic origin for the mafic and ultramafic rocks of the Salma eclogite association with a high degree of confidence. The normal mafic and Fe-Ti eclogites and piclogites also share geochemical features that indicate that their protoliths were indeed genetically related (Figs. 2.14 and 2.15). Relatively high Nb concentrations are consistent with a mantle-plume component having been involved in the petrogenesis of the protoliths. The combined geological, petrologic, geochemical, and zircon data show that the protolith of the Salma eclogite assemblage could have been a suite of predominantly plutonic mafic-ultramafic rocks like the third layer of the oceanic crust (layered gabbro)


82

Chapter 2

Figure 2.28. Plot of U/Pb age versus 176 Hf/177Hf showing a relatively narrow horizontal band of points that suggests the U-Pb system has been reset, where zircons with younger U-Pb ages have clearly been affected by nonzero Pb loss (ca. 1.9 Ga). TDM—model age. Figure is after Dokukina et al. (2014); reprinted with permission of Elsevier. DM— depleted mantle; CHUR—chondritic uniform reservoir.

in the Southwest Indian Ridge (mid-ocean ridge; see Dick et al., 2000, and references therein). An oceanic crustal affinity for the protoliths is confirmed by the abundant findings of relict mineral assemblages characteristic of early metamorphic to hydrothermal reworking in the ocean floor environment. These geochemical similarities between the Salma subduction-type eclogites and the Gridino eclogitized dikes naturally lead to speculation concerning the possibility that the two types of eclogites formed in interrelated geodynamic environments (Figs. 2.13–2.15; Mints et al., 2010b; Dokukina and Konilov, 2011). Evolution of Magmatism and Metamorphism Integration of the petrologic and geochronological data allows us to depict the history of the Salma eclogites in the form of a looplike P-T-t path (Fig. 2.23). The emplacement of melt and origin of the oceanic crust immediately followed its separation from the mantle source at 2.90–2.89 Ga. The proposed oceanic nature of the eclogitic protoliths allows us to consider the prograde evolution as a direct reflection of the P-T-t path of the subducting slab. The recorded metamorphic history of the oceanic plate involves transformation under oceanic floor conditions ca. 2.87 Ga; transition to high-pressure conditions as the plate submerged and eventually reached peak parameters (700– 775 °C and 13.0–14.3 kbar) 2.82 Ga; and retrograde evolution via stability field of granulite facies with an increase in temperature up to 730–780 °C ca. 2.72–2.70 Ga. At present, the Belomorian eclogite province is the first and the only natural example that makes it possible to reconstruct the transportation of mafic crust down to the realm of eclogite-facies conditions that existed in the Mesoarchean. The prograde branch of the P-T path started from the prehnite-pumpellyite facies, passing via a region of high temperatures and skirting the P-T range of blueschist-facies metamorphism. The Archean subduction recorded by the Salma association was significantly hotter than the known examples of

recent warm subduction (Carson et al., 1999; Peacock et al., 2002; Aoya et al., 2003; Page et al., 2003), and could be referred to as a hot subduction. At a depth of 25 km, where the Salma eclogites yielded temperature estimates at 640–670 °C, the Cascadia warm subduction zone, separating the Juan de Fuca and North America plates, has temperatures 100–200 °C lower, in the range of 450–550 °C. At shallow depths within the garnet-amphibolite field, the Salma P-T path crosses the basalt wet solidus line. This suggests the possibility of early partial melting of the subducting slab, before reaching eclogite-facies conditions. The metamorphic evolution of the Gridino rocks involves several Archean and Paleoproterozoic events or stages (Dokukina and Konilov, 2011; Dokukina et al., 2012, 2014). The U-Pb isotope studies of mafic and felsic metamorphic rocks of the Gridino area display a number of peaks in zircon ages (Fig. 2.29) related to successive events and different morphological types of zircons: (1) 3.0–2.9 Ga, corresponding to crystallization of the tonalitic protolith of the Belomorian gneisses; (2) 2.86–2.82 Ga, which date the emplacement of mafic intrusions, interaction between mafic magma and anatectic felsic melts, and their mutual crystallization; (3) 2.82–2.78 Ga, which date recrystallization under high-pressure (up to eclogite-facies) conditions; (4) 2.72–2.68 Ga, high-pressure granulite-facies metamorphism during decompression accompanied by anatectic granite formation; (5) 2.39 Ga, high-temperature thermal impact; and (6) 2.0–1.9 Ga, amphibolite-facies metamorphism. The P-T-t path of the metamorphic evolution of the Gridino rocks is shown in Figure 2.23. The prograde metamorphic evolution of the mafic dike swarm includes: (1) isobaric cooling of igneous rocks down to amphibolite-facies conditions (~5 kbar at ~600 °C); (2) burial with corresponding increase of pressure and temperature; (3) eclogitefacies metamorphism with peak conditions under 17–18 kbar (probably up to 22 kbar) and at ~800 °C (Fig. 2.23; Dokukina et al., 2012, 2014). This sequence of events is related to the retrograde



Figure 2.29. Integrated zircon age histograms for the Gridino eclogitized dikes constructed for concordant and near-concordant points with discordance of 0%–2%, after Dokukina et al. (2014). Reprinted with permission of Elsevier. MSWD—mean square of weighted deviates, (L)REE—(light) rare earth elements.

83


84

Chapter 2

metamorphism that included periods of cooling and decompression punctuated by thermal pulses caused by the emplacement of mantle plumes at 2.72–2.70, 2.5–2.4, and 2.0–1.9 Ga. As noted already, the last three events are characteristic of the early Precambrian history of the East European craton as a whole. Within the Belomorian eclogite province, they were accompanied by a corresponding temperature increase up to amphibolite- and granulite-facies conditions. An intervening period of cooling by 100–150 °C after 2.4 Ga has been deduced from the petrologic observations (Dokukina et al., 2014). We assume that similar cooling episodes preceded the plume impingement events ca. 2.72–2.70 and 2.0–1.9 Ga, although we have not yet been able to identify or constrain these events using petrologic evidence. The alternating periods of cooling-decompression and plume-related increase in temperature in the crust and underlying mantle are displayed in the P-T-t path (Fig. 2.23). The diagram also shows similar paths depicting retrograde evolution of the Salma eclogite association. Evidences for high-pressure metamorphism in the quartzfeldspathic rocks (granitoids and gneisses) are very scarce. A similar situation is common of other eclogite-bearing complexes (Rubie, 1986; Ernst et al., 1998; Gilotti et al., 2004). However, eclogite-facies mineral assemblages have been recently found in the granite leucosome crosscutting the metagabbro dike at Cape Vargas and in the enderbite vein in the northeastern Gridino Village. Granulite-facies mineral assemblages in felsic rocks have, however, been recognized more widely, typically manifested as garnet-kyanite gneisses and enderbite veins (Dokukina and Konilov, 2011; Dokukina et al., 2012, 2014).

Temporal Relationship between the Salma and Gridino Eclogites and Geodynamic Setting and Evolution of the Belomorian Eclogite Province Eclogites of both the Salma and Gridino associations are confined to and distributed throughout the migmatized TTG gneisses of the Keret complex, which plunges northeastward beneath the Inari-Kola microcontinent and lies structurally above the mafic-ultramafic association of the Central Belomorian suture zone (Fig. 2.10; Slabunov et al., 2006a; Mints et al., 2009, 2010b, 2014b). This compels us to suggest that the Keret eclogite-bearing TTG gneisses were originally derived from the Mesoarchean–Neoarchean active margin of the Inari-Kola microcontinent and the Kola continent. A striking feature of the eclogites is the general similarity of the P-T-t paths for both associations (Fig. 2.23; Mints et al., 2010b, 2014b; Konilov et al., 2011; Dokukina and Konilov, 2011). As was shown herein, the reconstructed protoliths suggested for the Salma eclogite were formed by intercalating gabbronorites, troctolites, and Fe-Ti gabbros resembling the third layer of oceanic crust formed at slow-spreading ridges. In turn, the magmatic source for the Gridino mafic dikes could have been a subducting midocean ridge. Geochronological data are integrated in Figure 2.30. A model of the overall evolution of the Belomorian eclogite province, which integrates the features described here, is presented in Figure 2.31. (1) The age of the oceanic-type protoliths of the Salma eclogites has been dated at ca. 2.9 Ga, although the extent of

Figure 2.30. Integrated histogram constructed on the basis of the geochronological data available, presented in Table 2.10.



85

Figure 2.31. A schematic model for the magmatic and metamorphic evolution of the Belomorian eclogite province, after Mints et al. (2014b). Reprinted with permission of Elsevier.


86

Chapter 2

oceanic lithosphere and relationship to other tectonic units remain unknown. (2) The precise timing and duration of the eclogite-facies metamorphism of the Salma assemblage have not been directly estimated. The reconstructed thermal regime of the inferred subduction zone is, however, consistent with subduction of a slow-spreading ridge (see following). On the other hand, the geochemistry of the eclogites from both the Salma and Gridino associations allows us to interpret the formation of the dike swarm beginning from 2.87 to 2.82 Ga in terms of injections of mafic magmas from a ridge source into the overlying crust of an active margin above a slab window. The pseudotachylytes studied on Izbnaya Luda Island provide evidence for seismogenic deformations inherent to recent subduction zones (Hyndman and Wang, 1993; Dragert et al., 1994; Wang et al., 2003). The time of this event probably fell within interval between the onset of spreading center plunging (ca. 2.9 Ga) and the initial emplacement of mafic dikes (2.87 Ga). Geochronological data and the geological relationships between dikes and veins are considered as well. It is reasonable to assume a close interaction between mafic intrusions and generation and segregation of felsic melts. Involvement of the lower crust in subduction between 2.82 and 2.78 Ga, which led to high-pressure metamorphism of Gridino dikes, may be explained by crustal delamination at the active margin and subsequent joint subduction of the lower crustal slices with mafic dikes and the oceanic plate (Fig. 2.31). With allowance for similarity in ages of the oceanic-type protolith of the Salma eclogites and the emplacement of the Gridino mafic dike swarm, which we have attributed to ridge subduction, and the complex interaction between crystallization-recrystallization processes in both dikes and crosscutting felsic veins under high-pressure conditions, it seems realistic to relate the time interval of 2.87–2.82 Ga to subduction of the oceanic lithosphere. It remains unclear, however, whether subduction terminated at this time. It cannot be ruled out that the second event dated by 2.82–2.78 Ga was related to the development of the Parandovo-Tiksheozero island arc and to the final collision that resulted in the amalgamation of Karelia, Kola, and Khetolambina continental blocks with formation of the Mesoarchean– Neoarchean Belomorian accretionary-collisional orogen (Mints et al., 2010b). The deformation associated with collision might have initiated the next high-pressure metamorphic event. (3) The Neoarchean–Paleoproterozoic history of the Belomorian eclogite province includes a series of events marked by crystallization and recrystallization of the zircon populations dated at 2.72–2.70 Ga, 2.39 Ga, ca. 1.9 Ga, and single grains 2.3–2.2 Ga in age. For comprehensive understanding of these dates, global and regional correlations must be considered. It is widely accepted that global peaks of U-Pb zircon age at 3.1–2.9, 2.8–2.7, 2.6–2.4, 2.2–2.1, 2.0–1.7, and 1.2–1.0 Ga, most of which are correlated with Lu-Hf, Sm-Nd, and Re-Os model ages, reflect juvenile crust production and related rapid crustal growth (Condie, 1998; Abbott and Isley, 2002; Hawkesworth and Kemp, 2006; Voice et al., 2011; Bradley, 2011, and references therein). The most prominent peak at 2.8–2.7 Ga suggests

an abrupt turning point in crust-forming processes in the early Neoarchean. The two predominant superplume eras in the Proterozoic occurred at 2.6–2.4 and 2.0–1.7 Ga (Abbott and Isley, 2002). These age peaks are thought to be associated with subduction of oceanic plates and suprasubduction magmatic processes (Condie et al., 2009a; Lee et al., 2011). However, it is difficult to imagine how global pulses of crustal growth might occur in response to conventional Wilson-style plate tectonics, particularly because there are no age peaks in the past billion years, from which we might surmise that subduction-related magmatism was the globally dominant process in the generation of new crust during certain periods of time. An alternative explanation is that periodicity in igneous activity reflects rapid crust formation during thermal pulses associated with the emplacement of mantle plumes (Hawkesworth and Kemp, 2006). (a) Both eclogite-bearing associations were one way or another uplifted to higher crustal levels characterized by a granulite-facies P-T regime by 2.72–2.70 Ga. The time interval of more than 60 m.y. between the second pulse of subduction and the granulite-facies metamorphism indicates that these events were independent of one another. At 2.7 Ga, a newly formed continental block of the eastern Fennoscandian Shield was an area of extensive sedimentation, magmatism, and high-temperature metamorphism related to the basically new stage of crustal evolution (chapter 3). These processes were confined to the ovalshaped Karelian-Belomorian and Kola provinces, reflecting influence of local mantle plumes (Mints et al., 2010b; Mints, 2014b). The Belomorian eclogite province occurred in the outer zone of the Karelian-Belomorian area, characterized by sporadic manifestation of granulite-facies metamorphism. We suggest that formation of granulite-facies mineral assemblages and specific (or related) zircon populations were linked directly to the aforementioned process. (b) Similarly, the thermal event ca. 2.4 Ga in age was accompanied by infiltration of metamorphic fluids, hydrothermal alteration, and crystallization of high-U and high-Th zircons. These processes were related to the early Paleoproterozoic pulse of mantle plume activity. This event, dated at 2.5–2.4 Ga, gave rise to the formation of a large igneous province (section 8.1.1) with gabbro-anorthosites, layered mafic-ultramafic complexes, charnockites, and bimodal volcanism developing throughout the eastern Fennoscandian Shield (Sharkov, 2006). This event also resulted in the creation of intracontinental basins characterized by intense sedimentation and granulite-facies metamorphism of the Archean basement rocks along with Paleoproterozoic fill of the basins. (c) Between 2.1 and 1.9 Ga, a relative lull was followed by reactivation of the mantle plume or, perhaps, by ascent of a new plume. The fill of sedimentary basins and their basement again underwent granulite-facies metamorphism. The episode of compression dated at 1.93–1.86 Ga led to the formation of synformal thrust-nappe assemblies in granulite-gneiss belts (sections 8.1.3 and 8.1.6). The rocks of the Belomorian province were reworked during that time. The Archean magmatic and metamorphic zircons were occasionally overgrown by colorless low-U rims dated


Mesoarchean Kola-Karelia continent at 1.9 Ga (Figs. 2.24E and 2.24F). Dating of similar zircons elsewhere in the Belomorian belt yielded the same age (Bibikova et al., 2004; Slabunov et al., 2006a; Mints et al., 2010b). Titanite from TTG gneisses dated at ca. 1.87 Ga (Bibikova et al., 2001) indicates equilibration at a temperature of ~700 °C at this time. The same Sm-Nd mineral–whole-rock isochron ages of metagabbronorite and metagabbro dikes (1.92–1.91 Ga) and Salma eclogites (1.89–1.87 Ga, with one exception at 1.79 Ga) show that late Paleoproterozoic metamorphism led to isotopic reequilibration of the Sm-Nd system in minerals. The Lu-Hf whole-rock–garnet–clinopyroxene isochrons for the Salmatype eclogites (Kuru-Vaara quarry and islands in the vicinity of Gridino Village) estimate the age in about the same range of 1.94–1.89 Ga (Herwartz et al., 2012). In contrast, the Sm-Nd ages of the Salma eclogites (ca. 3.0 Ga) and the Gridino mafic dikes (3.4–3.0 Ga) are quite robust (Mints et al., 2010b; Kaulina, 2010; Dokukina et al., 2010, 2012). The closing temperature of the Sm-Nd system in garnet estimated by Ganguly and Turone (2001) at 650–670 °C corresponds to amphibolite-facies conditions. Because little is known about diffusion of Hf in rock-forming minerals, it is not clear whether the Lu-Hf ages of eclogitic garnets reflect a closure temperature (TC) or the time of crystallization (Duchêne et al., 1997). It is known that the Lu-Hf closure temperature is higher than or equal to that of Sm-Nd in the same garnet. The actual TC values for Lu-Hf in garnet appear to depend largely on the size of the garnets and range from ~540 °C (0.24 mm radius) to higher than 700 °C (Scherer et al., 2000). The Lu-Hf and Sm-Nd ages obviously correspond to cooling rather than prograde garnet growth or peak metamorphic conditions (Kylander-Clark et al., 2007). The isotope geochemistry of eclogites has been used to estimate their cooling history (Mints et al., 2010b; Kaulina, 2010; Dokukina and Konilov, 2011; Dokukina et al., 2012). The cooling of the rocks to temperatures below 550–500 °C was recorded in the 40Ar/39Ar data on amphibole (Harrison, 1982). The U-Pb rutile age (1.79–1.76 Ga) is much younger and reflects the cooling time of metamorphic rocks to temperatures of 400–450 °C (Mezger et al., 1991). These values are confirmed by 40Ar/39Ar dating of muscovite and biotite in the leucosome at 1.79–1.69 Ga (Dokukina et al., 2010, 2012) for cooling of the system down to 425–280 °C (Harrison et al., 1985, 2009). (d) These data indicate that the Belomorian eclogite province associations were exhumed to middle or lower crustal depths ca. 1.77 Ga. Afterward, erosion episodes or younger tectonic events were responsible for final exhumation to the surface. Discussion on Existing Interpretations of the Available Geochronological Data There have been considerable discussions in recent years on the interpretation of geochronological data from the Belomorian eclogite province. Volodichev, Slabunov, Shchipansky, and their coauthors argued that two Archean (ca. 2.82 and ca. 2.72 Ga) and two or more Paleoproterozoic eclogite-facies events are recorded in the rocks of the Belomorian eclogite province (Volodichev et

87

al., 2004, 2005; Slabunov, 2008; Slabunov et al., 2011b; Shchipansky et al., 2012b). A quite different model was recently presented in a series of papers by Skublov et al. (2010a, 2010b, 2011a, 2011b), Herwartz et al. (2012), and Mel’nik et al. (2013). These researchers have made persistent attempts to “issue a final decision” on the late Paleoproterozoic “Svecofennian” (ca. 1.9 Ga) age of eclogitefacies metamorphism in the Belomorian eclogite province. The U-Pb zircon ages estimated by these authors are in agreement with the results that were previously obtained by other researchers. The ages of 2.88 and 2.70 Ga are interpreted by Skublov and others exclusively in terms of premetamorphic magmatic events, while the age estimates in the range of 1.91–1.84 Ga obtained from outer zircon rims with low concentrations of trace elements (“typical geochemical characteristics of zircons from eclogite” [Skublov et al., 2012, p. 443]) and Lu-Hf whole-rock–garnet– clinopyroxene isochrons were interpreted as corresponding to the late Paleoproterozoic age of the high-pressure metamorphism throughout the Belomorian eclogite province. In justifying their model, these authors ignored petrologic and geological evidences, focusing solely on the geochemical characteristics of the dated zircons. Similarity between the late Paleoproterozoic overgrowths and replacements of the earlier generations of zircons, on the one hand, and “typical eclogite” zircons defined on the basis of published data, on the other hand, is taken by these authors as a necessary and sufficient criterion for advocating a late Paleoproterozoic age of eclogite-facies metamorphism in the Belomorian eclogite province. However, individual zircon grains as well as rims surrounding older zircon cores, sharing similar geochemistry, morphology, and the same age, are widely distributed in various rock types throughout the Belomorian province, including rocks that have never undergone eclogite-facies metamorphism (Bibikova et al., 2004; Serebryakov et al., 2007; Kaulina, 2010). We also point out that in the paper by Skublov et al. (2009), which is concerned with dating of the Shueretskoe garnet deposit associated with late Paleoproterozoic pegmatite bodies, the authors made a conclusion that contradicts their own statement about unique and diagnostic attributes of “eclogite” zircons: “Zircons formed synchronously with garnet are similar to so-called eclogitic zircons in their geochemical characteristics (low concentrations of HREE and LREE, Th, and other minor elements), though the pressure of garnet deposit formation was significantly lower than the parameters of eclogite facies. Metasomatism resulting in the formation of garnet deposit occurred under conditions of a peak temperature of 650–680°C and a pressure of 7.8–8.5 kbar” (Skublov et al., 2009, p. 1547). It should be repeated that Skublov and his coauthors relied in their reasoning only on chemistry and isotopic parameters of zircon and garnet, completely ignoring geological and petrologic evidence of the sequence and evolution of metamorphism. Such reasoning probably could be applicable to reconstruction of a relatively short and simple metamorphic history no longer than a few tens of million years, e.g., from 93 to 42 Ma in eclogite from the Balma Unit, an ophiolite sheet on top of the


88

Chapter 2

Monte Rosa nappe in the Pennine Alps (Herwartz et al., 2008), or the metamorphic history of the Dabie-Sulu ultrahigh-pressure metamorphic rocks between 246–244 and 215–208 Ma (Schmidt et al., 2011), but not to multifold pulses of high-temperature metamorphic events under amphibolite- and granulite-facies conditions. Such an approach to the study of the Belomorian eclogite province rocks, with their complex multistage metamorphic history that lasted for about a billion years, would inevitably lead to erroneous interpretation of the crystallization conditions of zircon crystals, poorly reasoned suggestions on Lu distribution in garnet, and untenable conclusions about the metamorphic evolution and the age of eclogitic associations. Against this background, the reliable facts are especially important concerning (1) recrystallization of eclogites under conditions of granulite facies, which followed eclogite-facies metamorphism in Salma and Gridino ca. 2.7 Ga, and (2) intersection of the eclogitized dike by a metasomatic vein containing a population of 2.39 Ga zircons that are unique in composition and morphology. Role of the Belomorian Eclogite Province in Evolution of the Early Precambrian Crust in the Eastern Fennoscandian Shield For many years, it was assumed that the nature and precise location of the boundary between the Archean Kola and Belomorian provinces in the eastern Fennoscandian Shield were controlled by Paleoproterozoic structural elements (Kratz et al., 1978; Glebovitsky, 2005; Mints et al., 1996; Balagansky et al., 2006; Daly et al., 2006). However, recognition of the Belomorian eclogite province and the outcomes of its investigations make it possible to delineate and characterize this tectonic line as the Archean collisional boundary (Fig. 2.9). Conditions of Eclogite-Facies Metamorphism and Specificity of Subduction in the Belomorian Eclogite Province As shown here (Fig. 2.23), it is possible to identify eclogitefacies metamorphism in the Mesoarchean–Neoarchean Belomorian eclogite province as a product of hot subduction (Mints et al., 2010b; Konilov et al., 2011). Was such a thermal regime a specific characteristic of the Archean time? The features of thermal regimes in subduction zones, depending on the rate of spreading ridge approach, were highlighted by results of a two-dimensional numerical modeling by Uehara and Aoya (2005). In the absence of increasing mantle flow and shear heating, a ratio of the ridge approach rate u to subduction rate v was found to be an important parameter determining the thermal structure of the lithosphere just before ridge subduction. The P-T path of the Salma eclogite assemblage closely resembles the hottest one, which was calculated for the lowest u/v ratio (0.1) and corresponding to the lowest spreading rate (0.5–1.0 cm/yr). Bearing in mind the aforementioned similarity of the suggested Salma protolith assemblage to the gabbroic suite from the third layer of recent oceanic crust in the slow-spreading Southwest Indian Ridge, it seems quite reasonable to involve the slow-spreading ridge subduction to explain hot subduction in the Salma case.

2.4.5. Conclusions A summary of the main conclusions derived from this section includes the following points: (1) Eclogite-facies mafic rocks are hosted in gray gneisses of TTG affinity in the northeastern part of the Belomorian province (southern Kola Peninsula and northeastern Karelia). They are characterized by locally retained relics of omphacite and widespread exsolution textures of omphacite breakdown. Two types of eclogites are distinguished in the Belomorian province: (i) the subduction-type Salma association and (ii) eclogitized mafic dikes of the Gridino association. (2) Protoliths of the Salma eclogites are inferred to be alternating normal gabbro, Fe-Ti gabbro, and troctolites formed ca. 2.9 Ga in a slow-spreading ridge setting similar to that of the recent Southwest Indian Ridge. The main subduction event and eclogite-facies metamorphism of the Salma association took place in the time interval 2.87–2.82 Ga. (3) Mafic magma injections into the lower crust of the active margin and formation of the Gridino dike swarm were related to docking of the mid-ocean ridge to the subduction zone between 2.87 Ga and 2.82 Ga. Crustal delamination at the active margin and subsequent involvement of the lower crust in subduction between 2.82 and 2.78 Ga led to high-pressure metamorphism of Gridino dikes, culminating in eclogite-facies conditions during the collision event. This collision resulted in amalgamation of the Karelia, Kola, and Khetolambina blocks and the formation of the Mesoarchean Belomorian accretionary-collisional orogen. (4) Thermobarometry indicates a clockwise P-T path for Salma and Gridino associations with both crossing the granulitefacies P-T field. The succession of magmatic and metamorphic processes indicates a complicated posteclogite history with mantle plumes invoked to explain thermal reactivation and hydrothermal processes at 2.72–2.70 Ga, ca. 2.4 Ga, and ca. 1.9 Ga. The eclogite associations of the Belomorian province were exhumed to the middle and lower crust ca. 1.7 Ga, whereas erosion episodes or younger tectonic events were responsible for final exhumation at the present-day erosion surface. (5) These conclusions lead to the main statement: The Belomorian eclogite province provides evidence for subduction of the oceanic lithosphere and subsequent collision of continental blocks that occurred in the Mesoarchean around 2.90–2.78 Ga. (6) The thermal regime of the Archean eclogite-facies metamorphism represented by P-T parameters from the Salma and Gridino eclogite associations cannot be regarded as evidence for hotter Archean mantle as compared with later periods of Earth’s history. The high-temperature conditions during eclogite-facies metamorphism were related to subduction of a mid-ocean ridge.

MANUSCRIPT ACCEPTED BY THE SOCIETY 7 OCTOBER 2014 FIRST PUBLISHED ONLINE 29 JANUARY 2015 UPDATED MANUSCRIPT PUBLISHED ONLINE 4 MAY 2015 Printed in the USA


Geological Society of America Special Papers 2. Mesoarchean Kola-Karelia continent Michael V. Mints, Ksenia A. Dokukina, Alexander N. Konilov, et al. Geological Society of America Special Papers 2015;510; 15-88 , originally published onlineJanuary 29, 2015 doi:10.1130/2015.2510(02)

E-mail alerting services

click www.gsapubs.org/cgi/alerts to receive free e-mail alerts when new articles cite this article

Subscribe

click www.gsapubs.org/subscriptions to subscribe to Geological Society of America Special Papers

Permission request

click www.geosociety.org/pubs/copyrt.htm#gsa to contact GSA.

Copyright not claimed on content prepared wholly by U.S. government employees within scope of their employment. Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education and science. This file may not be posted to any Web site, but authors may post the abstracts only of their articles on their own or their organization's Web site providing the posting includes a reference to the article's full citation. GSA provides this and other forums for the presentation of diverse opinions and positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflect official positions of the Society.

Notes

© 2015 Geological Society of America