Anorthosite-Rapakivi Granite Petrogenesis in Connection with
Major Crustal Structures
EUG 10Convenors
Jean-Clair Duchesne
S. Bogdanova
M.D. Higgins
Anorthosite-Rapakivi Granite Petrogenesis in Connection with Major Crustal Structures
0287
In southern Norway, the post-collisional magmatism is represented by two suites: 1) the Rogaland anorthosite-mangerite-charnockite-granite (AMCG) suite emplaced between 930 and 920 Ma and 2) a suite of undeformed granitoids intruded in the vicinity of the Mandal-Ustaoset line; intrusion ages are poorly defined although they cannot be older than 980 and younger than 890 Ma. Petrographical, geochemical and isotopic data show that these two suites are clearly different. The most conspicuous difference is the abundance of hydrous phases (biotite, hornblende) in the granitoid suite and their quasi absence in the AMCG suite. The suites also differ by their SiO2 ranges: 59% to 78% for the granitoid suite and 49.4% to 73% for the AMCG suite. Moreover, the AMCG suite is a high Fe#, A-type suite (Duchesne & Wilmart, 1997), whereas the granitoid suite is high-K calc-alkaline, with higher CaO and MgO contents. In ORG-normalised spidergrams, Th displays a negative anomaly in the AMCG suite, which persists throughout differentiation (Vander Auwera et al., 1998), whereas there is no Th anomaly in the granitoid suite. Also in the granitoid suite, Ba displays no anomaly or a negative one. This reflects progressive subtraction of biotite and K-feldspar in the crystallizing assemblage. In the AMCG suite, there is a positive Ba anomaly reflecting the absence of these latter phases in the cumulates. Both suites show overlapping ranges in initial Sr isotopic composition (AMCG suite: 0.7029 - 0.7085; granitoid suite : 0.7027 - 0.7056 at 930 Ma; this study) but strongly differ in their Nd isotopic composition (AMCG suite: <epsilon> Nd = 0 to +5.7; granitoid suite: <epsilon> Nd = -0.85 to -4.90) (Demaiffe et al, 1986; Menuge, 1988; Nielsen et al., 1996; this study). These data suggest that, despite their contemporaneity, the parent magmas of the two suites probably result from the partial melting of two different sources: a LILE-depleted source (either the upper mantle or a mafic rock derived from it) for AMCG and an undepleted or slightly enriched source for the granitoids. Moreover, phase equilibrium data bring additional constraints and show that the parent magma of the anorthosites result from the partial melting of a lower crustal anhydrous gabbronoritic source (Longhi et al., in press). The source of the granitoids is still to be determined.
Duchesne JC & Wilmart E, J. Petrol., 38, 337-369, (1997).
Vander Auwera J, Longhi J & Duchesne JC, J. Petrol., 39, 439-468, (1998).
Demaiffe D, Weis D, Michot J & Duchesne JC, Chem. Geol., 57, 167-179, (1986).
Menuge JF, Contrib. Mineral Petrol., 98, 363-373, (1988).
Nielsen FM, Campbell IH, McCulloch M & Wilson JR, J. Petrol, 37, 171-193, (1996).
Longhi J, Vander Auwera J, Fram M & Duchesne JC, J. Petrol, (in press).
0312
The Fennoscandian rapakivi complexes consist of a number of batholiths and smaller plutons, which are located from Karelia in Russia in the east to central Sweden in the west. The U-Pb ages of the rapakivi rocks indicate a gradual decrease in age from east to west, except for the most eastern complex, the Salmi batholith. Palaeomagnetic data excists for a number of these complexes and related dykes and with this study we want to put the formation of the rapakivis into a plate tectonic context. The calculated pole positions do not reveal any clear trends, however, poles calculated from rapakivis in Finland are located at somewhat lower latitudes and more eastern longitudes than poles of corresponding rocks in Sweden. This probably reflects an APW related to the general age difference between the rapakivi complexes. The palaeolatitudes of Fennoscandia at the time of the intrusion of the rapakivis are restricted to a fairly narrow latitudinal range between 0° and 27° northern latitudes. In spite of poorly constrained ages of magnetization and sometimes large error bars in the calculated palaeolatitudes there seems to be a slight trend of increasing palaeolatitudes with decreasing age. A similar trend of changing palaeolatitudes has also been observed during the intrusion of anorthosite-rapakivi complexes in the Ukrainian Shield. Possible origin of the rapakivi intrusions will be discussed with reference to these palaeomagnetic results.
2994
Two major domains, the West Lithuanian Granulite (WLG) Domain and East Lithuanian Belt (ELB), separated by the Middle Lithuania Suture Zone (MLSZ) were distinguished in the concealed crystalline basement of Lithuania. Numerous granitoid and mafic intrusions feature along the major or minor boundaries and tectonic zones.
Granitoids which might be referred to as anorthosite-mangerite-charnockite-rapakivi granite (AMCG) magmatic suite have been recognized in southwesternmost part of the WLG, in the NW-SE trending Nemunas Fault Zone (e.g. Rukai, Vabalai, Usenai, etc). The rocks consisting of pyroxenes, biotite, plagioclase, perthite, quartz, zircon, apatite, Fe-Ti oxides with minor carbonates. In the central part of the WLG some of the massive or shared K-feldspar granitoids with related basic rocks and anorthosite veins (e.g. Geluva, Pamituvys, Pramedziuva, etc.) are situated along an E-W trending shear zone.
In the southern part of the MLSZ, the Lazdijai area, numerous bodies of porphyritic hornblende and hornblende-biotite granitoids are an extension of the Mazury granitoids in Poland having an age of 1.5 Ga (Claesson, 1996). The Mazury granitoids including those of the Lazdijai area and the Suwalki anorthosite complex might form a composite AMCG suite.
In the ELB, some of the porphyritic and equigranular K-feldspar granites with biotite having A-type and I-type characteristics and relevant rare earth element distribution are regarded as anorogenic granites. The 1.5 Ga (Sundblad et al. 1994) Kabeliai granites plot mostly in the within-plate granite fields. Synchronous porphyri type Cu-MO mineralisation is related to the Kabeliai granites.
Thus, high level rapakivi granites (the ELB) and deeper-level mangerite- and charnockite-dominant suite (the WLG) are distinquishable in the crystalline basement of Lithuania. Those rocks tend to be mainly associated with the major boundaries or E-W trending shear zones. The described suites appear to be intruded within 150 myr or less of the end of the last orogeny (ca. 1.7-1.66 Ga, Bogdanova et al., 1996).
Bogdanova SV, Page LM, Skridlaite G, Taran L, GFF, 118, 10 - 11, (1996).
Claesson S, Conf. Abs, 11, (1996).
Sundblad K, Mansfeld J, Motuza G, Ahl M & Claesson S, Mineralogy and Petrology, 50, 43 - 57, (1994).
3145
The Suwalki Anorthosite Massif (SAM) is one of three separated anorthosite intrusions (Sejny and Kêtrzyn) emplaced in a Proterozoic E-W thrust zone in NE Poland. Anorthosites are surrounded by norites, gabbronorites and diorites, and enclosed in rapakivi granites. The complex is covered by 700-1000 m-thick Phanerozoic sedimentary rocks.
SAM is a massif-type anorthosite, showing a polybaric evolution, characterized (1) by megacrysts of iridescent plagioclase (ca. An55) with high Sr content (800 -900 ppm), (2) by kinked and polygonized megacrysts of orthopyroxene with plagioclase oxy-exsolutions, (3) by Fe, Ti, P-rich jotunitic chilled melts (Duchesne et al. 1998). Fe-Ti-V oxide deposits with subordinate Fe, Cu, Ni and Co sulfides are conspicuous.
Re-Os data (Stein et al. 1998) for sulfides and magnetite yield two parallel isochrons with ages of 1559 ± 37 Ma (Jezioro Okragle and Krzemianka) and 1556 ± 94 Ma (Udryn), with correspondingly very high initial 187Os/188Os ratios of 1.16 ± 0.06 and 0.87 ± 0.20. The data indicate a source that is highly crustal and significantly older for the Suwalki ores and associated anorthosite. The isochron ages are in good agreement with U-Pb zircon ages of ca. 1500 Ma obtained on the Mazury rapakivi complex (Claesson et al. 1995).
Initia <epsilon> Nd and <epsilon> Sr values, calculated for 1.5 Ga, are -5.3 to +1.6 and 30 to 42 respectively. This confirms that the source of the SAM rocks includes older crustal material. Nd TDM model ages for different Suwalki samples, with Nd concentrations from 2 to 125 ppm, vary between 1.7 and 2.3 Ga. There is no correlation between Nd concentration and model age, but the 147Sm/144Nd ratios correlate positively with the model ages. Samples with low 147Sm/144Nd of <0.09 have model ages down to 1.7 Ga, while the model ages for samples with 147Sm/144Nd >0.09 typically are 2.0 Ga or more. The latter are similar to model ages determined for the adjacent Mazury rapakivi granite and also to values for various rock types from the Svecofennian Domain to the north and west. These results are consistent with formation of Suwalki magmas by melting of mafic rocks with a crustal pre-history of several hundred Ma.
In view of recent phase diagram constraints on the origin of anorthosites (Longhi et al. in press), these results point to a possible origin of SAM by melting of gabbronoritic crustal rocks.
Longhi J, Vander Auwera J, Fram M & Duchesne JC, J. Petrology, in press
Duchesne JC, Vander Auwera J & Wiszniewska J, EUROBRIDGE workshop, Tallinn, (1998).
Stein HJ, Morgan JW, Markey RJ & Wiszniewska J, Geophysical J, 20, 111-114, (1998).
Claesson S, Sundblad K, Ryka W & Motuza G, Terra Nova, 7, 107, (1995).
2151
The extensive area of rapakivi granites and associated mafic dykes in SE Finland are thought to be associated with rifting in the mid-Proterozoic. Whether the magmatism and rifting were synchronous is not known. An extensional regime was advocated as the extensive mafic dyke swarms associated with the granites lie within a region of thinner crust. These associated Fe-rich tholeiites are concentrated in two swarms. The larger of the two swarms (Häme) extends for some 250 km in a WNW - ESE direction from the NW corner of the Wiborg Batholith; The smaller Suomenniemi swarm lies to the east of the Häme dykes and is slightly arcuate in plan. It trends more to the north than the Hame swarm, extending for roughly 110 km. The dykes appear to have been intruded in three main pulses: at 1665, 1645 and 1635 Ma. A similar, yet marginally younger, array of dates has been found for the emplacement of granites in the Wiborg Batholith. Concomitant with the last pulse of diabases at 1635 Ma was the intrusion of quartz feldspar porphyry (QFP) dykes, concentrated near the margins of the batholith. In at least four instances the QFP and diabases form composite dykes that show mixing as well as mingling of silicic and mafic magmas, giving rise to relatively minor amounts of hybrid rocks associated with the Suomenniemi Batholith - a satellite of the Wiborg Batholith. The four composite dykes discussed here are divided into two categories, in which: (I) mingling of the silicic and basic components (with slightly different Nd isotopic signatures) is prevalent but some (indeterminant) amount of mixing has occurred: Korpijärvi and Kuusenhako dykes; (II) mixing is prevalent but some remnants of mingling are present in varying degrees: Kirkkovuori and Niemenmaa dykes. Nd isotopic data do not, in general, tell the two categories apart; their initial epsilon-Nd (at 1635 Ma) values range from -1.1 to -0.4 and appear to be largely controlled by fractionation stage of the mafic magmas.
0721
Anorthosite-rapakivi granite complexes (ARGC) 1.8-1.5 Ga in age are a specific magmatic associations of the Proterozoic. The classical location of such complexes is the western part of the East-European Craton. They occur mainly within the Svecofennian domain, which was formed after closure of the Svecofennian ocean, and partly in the Ukrainian Shield. The feature of the western part of the craton is unusually large thickness of crust reached in some places 50-60 km. The ARGC are absent to the east from the Svecofennian domain, where Archaean crust of normal thickness (about 40 km) predominated. Magmatism, simultaneous with the ARGC, was rather rare there. The ARGCs are large multistage batoliths which formed during 20-25 Ma. They are composed mainly by anorthosites and granites where the latter predominate. In certain portions, they look like as coarse-layered intrusions with graduate transition from anorthosites through norites, monzonites and diorites to granites (Salmi, Ahvenisto, Korosten, etc. massifs). According to geochemical and isotopic data, primary magmas of rapakivi derived from crustal material and anorthosites - from mantle-derived melts, contaminated by lower-crustal material. Diabase dykes are closely associated with anorthosite-rapakivi granite complexes; they often intruded during the formation of the ARGCs, as shown by cross-cutting relationships with granites and by mingling. These dykes originated from Fe-Ti basalts, demonstrating that formation of anorthosite-rapakivi granite complexes occurred simultaneously with melting of asthenospheric mantle beneath them. The ARGCs on their tectonic setting were within-plate formations, which appeared above hot spots in condition of very thick (60-80 km in the moment) continental crust. The latter was reworked during hot-spots activity and mantle highs occurred beneath the ARGC magmatic systems. On their geochemical features rapakivi rather close to A-granites, especially rare-metals Li-F granites and ongonites. The latter situated in complicated geodynamic environments, where oceanic spreading zones were overrided by continental plate (early Mesozoic of the Central Asia, and Cenozoic of the Western USA). Probably, existence of a former spreading center beneath the thick continental crust was important for generation of such types of magmatic activity.
2016
The Lac-St-Jean Anorthositic suite (AMCG) (1160-1140 Ma) consists of multiple injections of magma which together form a huge mass covering 20 000 km2. Its emplacement was controlled by two pre-existing shear zone systems. The NE-SW system comprises three major sub-parallel shear zones, each of which is 200 to 300 km long (St-Fulgence, Pipmuacan, Chute des Passes). The second system is composed of several NNE-SSW shear zones (Péribonka river, Bégin, ans Chabot megadykes). The NE-SW shear zones splits the anorthositic mass into two domains, each with several lobes. Each lobe has its own temporal, compositional, textural, structural and metamorphic characteristics. The Lac-St-Jean Anorhtositic suite was emplaced along these major structures. Some of these were reactivated during the Grenvillian orogeny which produced thrusting of the different lobes towards the NW.
0264
AMCG (Anorthosite, Mangerite, Charnockite, Granite) complexes, including Rapakivi granites, are classicaly considered as anorogenic. The consensus model by crustal underplating involves a deep-seated magma chamber in which differentiation of a mantle-derived magma leads to a floated cumulate at the roof. Diapiric uprise of anorthosite lumps through an unstressed lower crust produces massive anorthosites. It has been confirmed that the model is viable from a rheological stand point (Barnichon et al. in press). It may however be anticipated that weakness zones in the crust would favour emplacement by channeling diapirs and other related magmas at low energy costs. Therefore correlation of anorthosites with terranes boundaries (e.g. Nain, Laramie) or lineaments (e.g. Lac St-Jean) would not be coincidental. Recent petrological data (Longhi et al. in press) have shown that the parent magmas pooling in the deep-seated magma chamber (from high alumina basalt to jotunite) cannot be produced by mixing of mantle-derived melts with crustal material, but result from melting of gabbronorite in dry conditions. Crustal tongues produced by underthrusting the lower crust in collision zones - e.g. Sveconorwegian orogeny (Andersson et al. 1996) - can be brought to the necessary high PT conditions after thermal relaxation to melt gabbronorites from the crust (Duchesne et al. 1998). This model relates AMCG complexes to active or reactivated deep-crustal structures. In the EUROBRIDGE transect, the 1.5 Ga Suwalki anorthosite, of jotunite derivation (Duchesne et al. 1998) and with a highly crustal Re-Os signature (Stein et al. 1998), emplaced with Rapakivi granites along an E-W lineament.
Andersson M, Lye JE, Husebye ES, Terra Nova, 8, 558-566, (1996).
Barnichon JD, Havenith H, Hoffer B, Charlier R, Jongmans D, Duchesne JC, Tectonophysics, in press
Duchesne JC, Vander Auwera J, Liégeois JP, Longhi J, Geophysical J, 20, 70-71, (1998).
Duchesne JC, Vander Auwera J, Wiszniewska J, EUROBRIDGE workshop, Tallinn, (1998).
Longhi J, Vander Auwera J, Fram M, Duchesne JC, J. Petrology, in press
Stein H, Morgan JW, Markey RJ, Wiszniewska J, Geophysical J, 20, 111-114, (1998).
2666
The Fennoscandian rapakivi granites were intruded in the Subjotnian (1500-1650 Ma) within the Svecofennian Southern Finland -Central Sweden Island Arc complex in an extensional environment that developed 200 Ma after the Svecofennian orogeny (1800-1900). The thick Svecofennian crust (55-65 km) has large ovoid thinnings (40-45 km) associated with the outcropping rapakivi granites. Most of the rapakivi granites were intruded within or at the borders of the northern Baltic Sea, including Bothnian Sea, Gulf of Finland and Gulf of Riga. BABEL reflection profiles 1,6,7, and C image well crustal structures associated with the large rapakivi batholiths in the Bothnian Sea, Åland rapakivi in the south and Bothnian rapakivi in the north. The rapakivi granites are imaged as unreflective blocks delineated by listric and normal faults. The listric shear zones may be inverted Svecofennian thrust zones. High reflective structures within the batholiths are related to contacts between mafic dykes or intrusions and rapakivi granites. The highly reflective lower crust, that thus is composed of two components, shows upward concave structures that have resulted from the uprising movement of the lower crust. The high reflectivity is interpreted as being caused by mafic underplating and intraplating, that acted as thermal energy source the lower crustal bath melting resulting in the rapakivi magma generation. The listric shear zones that detach either at the lower crust to middle crust boundary or the Moho boundary, provide pathways from the lower crust to the upper crust for the rapakivitic and mafic magmas. Space created by the extending upper crust is occupied by both the rapakivi granite magmas and the uprising lower crust. The cooling underplated material crystallises as mafic unreflective lowermost crust and a new Moho boundary develops. In the cooling phase, also cauldron subsidence structures in the rapakivi granites may develop, as suggested by the reflection patterns in the Bothnian rapakivi batholith.
3338
The western part of the East European Craton (EEC) is known as a region of voluminous anorthosite-rapakivi granite magmatism at 1.65-1.45 Ga. New, combined geophysical-geological research in that part of the EEC including the area between the Baltic and Ukrainian Shields also demonstrates the presence of several craton-scale, E-W trending shear zones. These are superimposed onto the Palaeoproterozoic crustal-tectonic patterns and displace the boundaries of earlier-formed rock units. The largest of these zones are wide packages of subparallel faults. They often exhibit dextral strike-slip and control the distribution of anorthosite-rapakivi plutons and dykes associated with these plutons. The dyke swarms suggest NW- and SW-trending extensional faulting but also follow some of the E-W faults. Ar/Ar dating of amphiboles from mylonites along some of the shear zones demonstrates concomitant anorogenic magmatism and fault-tectonic activity (e.g. Bogdanova et al., 1996).
The E-W shear zones are seen well in the gravity and magnetic fields as well as the seismic structure of the lithosphere. They coincide with zones of crustal thinning and high reflectivity, and with Moho irregularities (e.g. Korja et al., 1993). According to the BABEL profiling, many of the shear zones in the Baltic Shield are listric. They were presumably developed in simple-shear extensional settings (Korja and Heikkinen, 1995). The formation of the E-W striking shear zones may have been a distant, transcurrent-faulting consequence of convergent tectonics along the western margin of the Svecofennian domain of the Baltic Shield (cf. Åhäll & Gower, 1997).
Transtensional models explain best the spatial and age relationships between the Mesoproterozoic NW- and NE-striking extensional structures, the E-W trending shear zones, and the anorogenic magmatism. In this case, the NW- and NE-striking faults may represent bends or stepovers which followed older, Svecofenian crustal discontinuities. The E-W shear zones are interpreted as major dextral strike-slip faults. The transtensional model also accounts for the general westward younging of the anorthosite-rapakivi magmatism in the region, which is combined with local eastward younging in some of the complex anorogenic intrusions, e.g. the Ragunda massif in central Sweden (Persson, 1996) and the Salmi massif in Karelia. Furthermore, it explains the semi-coeval formation of the granitoids and the related dyke swarms, and the variation of the strikes of the dykes observed in the field.
Abramovitz T, Berthelsen A & Thybo H, Tectonophysics, 270, 259-270, (1997).
Åhäll K-I & Gower Ch F, GFF, 119, 181-191, (1997).
Bogdanova SV, Page LM, Skridlaite G & Taran LN, GFF, 118, A10-A11, (1996).
Korja A & Heikkinen P, Tectonics, 14, 504-517, (1995).
Korja A, Korja T, Luosto U & Heikkinen P, Tectonophysics, 219, 129-152, (1993).
Persson AI, 7th Int. Symp. on Rapakivi Granites and Related Rocks, Helsinki, Abs, 57, (1996).
1665
The ~1.57 Ga anorogenic rocks in SW Finland comprise granites and ignimbrites of the Åland rapakivi batolith, tholeiitic dyke swarms and anorthosites. Reported ages for the felsic and mafic rocks overlap and are consistent with field observations of simultaneous intrusion; e.g. rapakivi granites intermingled with mafic enclaves carrying anorthosite fragments, and areas of tholeiite - rapakivi mixtures.
In the mafic rock association, the most primitive tholeiites show Nd and Sr isotopic compositions of a mildly depleted mantle, while the composition of more evolved tholeiites is consistent with the activity of AFC- processes during differentiation. Plausible contaminants are crustal melts similar in composition to the rapakivi granites and related quartz-feldspar porphyries. Isotopic and other geochemical data also indicate strong crustal influence during anorthosite petrogenesis.
P-T data indicate that part of the labradorite megacrysts in the anorthosites crystallized in magmachambers at 5-7 kbars, the same depth where deep seated mineral assemblages of the rapakivi granites were formed prior to ascent to subvolcanic levels. The time when the cores for rapakivi megacrysts nucleated is poorly constrained, but available mineralogical and Sm-Nd, Rb-Sr, U-Pb isotope data on the inherited components in feldspars and zircons indicate that a "protorapakivi" protholith already existed prior to the anorogenic magmatic event. Attempted thermobarometrical assessments evidence midcrustal levels for generation of rapakivi granite magmas (5-7 kbars).
The transportation of the rapakivi granite magma to subvolcanic levels took place in a mode of subisothermal ascent (whereupon the rapakivi texture was formed), the initial temperature was around 780°C. The magmas solidified at approx. 0.5-2.5 kbars and 680-720°C. Occasionally hot mafic magmas invaded shallow-seated granitic magmachambers, superheated the crystal saturated rapakivi magmas up to 950°C causing extrusion and deformation of flow-foliated ignimbrites.
3304
Experimental data in the range of 1 bar to 13 kb enable us to map the liquidus equilibria relevant to Proterozoic (massif) anorthosites and related mafic rocks. Massif anorthosites are widely believed to have formed by accumulation of plagioclase into high-Al basaltic liquids. Mantle-derived basaltic liquids, fractionating at pressures sufficiently high (10-13 kb) to crystallize the highly aluminous orthopyroxene megacrysts typically observed in anorthosite massifs, reach plagioclase saturation at low normative silica contents. Peritectic-like equilibria (e.g., liq + opx = plag + cpx + sp) and a thermal divide on the plagioclase + pyroxene liquidus surface ensure that mantle-derived liquids become nepheline normative with further crystallization and crustal assimilation at depth. Such liquids cannot produce the full range of troctolitic/noritic to troctolitic/gabbroic mineral assemblages observed in anorthosite massifs without extensive low-pressure granite assimilation. Conversely, the array of plausible anorthosite parental liquids not only lies along the trace of the plagioclase + 2-pyroxene cotectic from 10 to 13 kb, but also straddles the thermal divide on the plagioclase + pyroxene liquidus surface. This condition requires mafic source regions, such as lower continental crust or foundered mafic plutons, for liquids parental to massif anorthosites and associated mafic intrusions.
3434
Rhenium and osmium abundances and Os isotopic compositions were measured for nine sulfides and four titanomagnetites from the Suwalki anorthosite massif in extreme northeast Poland. The titanomagnetites are an order of magnitude lower in their Re (0.4-1.5 ppb) and Os (0.036-0.144 ppb) concentrations relative to co-precipitating pyrrhotite, pyrite, and chalcopyrite which yield very consistent concentrations for Re (30-55 ppb) and Os (1-6 ppb). Parallel lines with slopes of ~1 for sample-paired titanomagnetite and sulfides on Re versus common Os concentration plots indicate that both Re and Os behave similarly during crystallization in their high preference for any sulfide phase over magnetite.
Three deposits within the anorthosite massif were analyzed, and two different but highly elevated initial 187Os/188Os ratios emerged. An age of 1559 ± 37 Ma (n = 9) with an initial 187Os/188Os of 1.16 ± 0.06 for the Jezioro Okragle and Krzemianka deposits is essentially identical to an age of 1556 ± 94 Ma (n = 3) for the Udryn deposit (Stein et al., 1998; Morgan et al., submitted). Udryn, however, yielded a marginally lower initial 187Os/188Os of 0.87 ± 0.20. The high initial 187Os/188Os combined with the Proterozoic Re-Os age require that the source for Suwalki anorthosite is crustal and significantly older, possibly involving Archean rocks. Assuming an average crustal value of 50 for 187Re/188Os and using the 1559 Ma age and initial 187Os/188Os of 1.16 yields a 2777 Ma age for Suwalki source rocks. Widespread Phanerozoic cover has severely limited knowledge of basement rocks in Poland, and no Archean has been recognized in the Suwalki region (Bogdanova and Gorbatschev, 1998). Selection of 187Re/188Os higher than average continental crust, reflecting an increasing component of mafic crust in the source, moves the source age for Suwalki anorthosite and ores toward younger values that include the Proterozoic. Based on 2.0 Ga TDM Nd model ages for the Suwalki source (S. Claesson, pers. comm.), the 187Re/188Os in that source would be ~140, a reasonable value for crustal rocks with a mafic component. Regardless of Archean or Proterozoic source age, the high initial 187Os/188Os derived from the Re-Os isochron data require that the source for the anorthosite massif and its oxide-sulfide ores is crustal, and not mantle-derived.
The Suwalki anorthosite massif provides a real geologic example of the crustal tongue melting model of Duchesne et al. (1998), a model based on experimental data that constrain parental magma composition and source for anorthosites (Longhi et al., in press). The crustal tongue model suggests that slabs of dry mafic lower crust, downthrusted along flattened terrane boundaries, are melted to provide the parental magma for anorthosites. This may be the only existing model to explain the high initial 187Os/188Os ratios derived from the Re-Os isochron data.
Bogdanova SV & Gorbatschev R, Geophysical Journal, 4, 20, 60-63, (1998).
Duchesne JC, Liegeois JP, Vander Auwera J, & Longhi J, GAC-MAC Abstract Volume, 23, A47-48, (1998).
Longhi J, Vander Auwera J, Fram MS, & Duchesne JC, Journal of Petrology, (in press).
Morgan JW, Stein HJ, Hannah JL, Markey RJ & Wiszniewska J, Mineralium Deposita, (submitted).
Stein HJ, Morgan JW, Markey RJ & Wiszniewska J, Geophysical Journal, 4, 20, 111-114, (1998).
2615
Nd isotopic data indicate that the suture between the Archean and Proterozoic in the Grenville Province is commonly exposed near to the parautochthon / allochthon boundary. However, deep seismic results (Lithoprobe project) suggest that it continues to the SE beneath the allochthonous thrust sheets emplaced during the Grenville orogeny. The Massif-type Anorthosite Complexes (MAC) were emplaced into the allochthon during the Mid-Proterozoic, but must have been rooted much closer to the Archean / Proterozoic boundary before thrusting. The association of MAC with the Archean / Proterozoic boundary is seen elsewhere (Laramie, Nain, etc.) and may have a thermal origin: The large size of the crystals in MAC, their size distribution and their commonly uniform composition indicate that they were formed by extreme textural coarsening (Higgins, 1998). Such a process requires that the magma was near the plagioclase liquidus for a long period. MAC could not form during the Archean because the heat-flow was too high and magmas did not pond in the crust for sufficient time. Although the textures of Archean Anorthosite indicate that some textural coarsening did occur. During the Proterozoic, the Archean crust and lithosphere may have acted as a thermal blanket, deflecting heat and magma from the mantle towards the edge of the craton, where belts, such as the Grenville, were forming. Even at this time the parental magmas of the MAC could only rise to the mid-crust, where they are now exposed, along second order sutures (major shear zones). In the Phanerozoic the flux of heat was generally too low for textural coarsening to occur, except in very large intrusions (e.g. Sept Iles, Canada).
Higgins MD, J Petrology, 39, 1307-1325, (1998).
1111
The Adrar des Iforas (SW Tuareg shield) alkaline province in Mali (560-540 Ma) comprises 15 granitic ring complexes up to 30 km in diameter, rhyolitic dykes tens of km long and remnants of rhyolitic plateaux 2-3 km thick which probably covered the major part of the region (60000 km2). It was emplaced in a short time interval (<20 Ma) in a post-collisional setting (collision at c. 630 Ma) when stopped the horizontal movements of terranes linked to the Pan-African amalgamation in the Tuareg shield (Black et al., 1994). Alkaline magmatism occurred generally close to rheologically contrasted terrane boundaries during late movements along major shear zones related to reversals in the stress field. Sr, Nd, Pb isotopes favour a deep mantle source (asthenosphere-lithosphere interface), distinct from the preceding high-K calc-alkaline source (Liégeois et al., 1998). Geochemically, a strong "plagioclase effect" is evidenced; the gigantism of this province suggests in consequence the existence of massif-type anorthosites at depth.
Presence of massif-type anorthosites associated with granitic alkaline ring complexes is known at c. 410 Ma in Aïr (Niger; Demaiffe et al., 1991). Clearly anorogenic, this province followed the cessation of the Pan-African movements by >100 Ma; however, it was emplaced during the reactivation of a Pan-African mega-shear zone. These ring complexes share strong similarities, including source characteristics, with the post-collisional Iforas alkaline rocks.
In Rogaland-Vest Agder (Norway), the major regional granulite/amphibolite metamorphism occurred at 1024-970 Ma. Anorthosite-mangerite-charnockite-granite group (AMCG) intruded at 930 Ma and forms an alkaline series (Duchesne & Wilmart, 1997) subsequent to high-K calc-alkaline magmatism. AMCG emplacement occurred during the final movements of the adjacent Feda shear zone that probably brought these deep-seated intrusions near the surface. The lithospheric scale of the Feda shear zone is attested by geophysical data showing an important Moho offset at depth (Andersson et al., 1996). The post-collisional Rogaland massif anorthosites are very similar to anorthosites from anorogenic settings.
Alkaline granitic ring complex and anorthosite provinces have then strong similarities although representing different structural levels (Bonin, 1996): they are alkaline with similar trace element enrichment (sliding normalization; Liégeois et al., 1998), they have a juvenile source (OIB-type mantle or young mafic lower crust derived from it), they are located along major shear zones marking either their late movements following major collision or their intraplate reactivation while tapping the same source, they are voluminous although short-lived and localized to some terrane boundaries rheologically well-defined.
Andersson M, Lie JE & Husebye ES, Terra Nova, 8, 558-566, (1996).
Black R, Latouche L, Liégeois JP, Caby R & Bertrand JM, Geology, 22, 641-644, (1994).
Bonin B, Petrology and geochemistry of magmatic suite of rocks in the continental and oceanic crusts (Demaiffe D, editor), ULB-MRAC, Bruxelles, 201-217, (1996).
Demaiffe D, Moreau C, Brown WL & Weis D, Earth Planet Sci. Lett, 105, 28-46, (1991).
Duchesne JC & Wilmart E, J Petrol, 38, 337-369, (1997).
Liégeois JP, Navez J, Hertogen J & Black R, Lithos, 45, 1-28, (1998).
3165
The relative importance of terrane boundaries in the origin of Proterozoic anorthosite complexes (PAC) is currently being evaluated. This is due in large part to the discovery of: (1) multiple episodes of anorthositic magmatism separated by several 100 myr along terrane boundaries in the Laramie and Nain complexes (Scoates and Chamberlain, 1996; Hamilton et al., 1998), and (2) the Voisey's Bay massive Ni-Cu-Co deposit, hosted within troctolitic dikes of the Nain Plutonic Suite, and emplaced along the suture zone between Paleoproterozoic and Archean basement. Both the Nain and Laramie complexes also contain important volumes of troctolitic rocks (plag+ol±cpx/mt/ilm) which could be appropriate parental magmas for some of the olivine-bearing anorthosites.
Troctolitic rocks in PAC occur either as cumulates in layered (Kiglapait) or massive intrusions, or as high-Al,Fe dikes (Laramie, Adirondacks) and marginal intrusions (Harp Lake) with near-melt compositions. Troctolitic magmas have the least evolved compositions in PAC with Mg#<0.7 and An<60, and high alumina contents (15-19 wt%) coupled with positive Eu anomalies which do not reflect plagioclase accumulation, but result primarily from enrichment in plagioclase components during decompression melting of suspended plagioclase. All troctolitic magmas are depleted in Ni-Cu-Co compared to basaltic magmas (Ni<300 ppm; Cu<100 ppm) and S contents are in the range of 200-500 ppm for the highest MgO magmas. They are significantly S-undersaturated, thus an external source of S or bulk crustal contamination for a Voisey's Bay-type deposit is necessary. Due to very low Nd concentrations, the Nd-Sr isotopic systematics of troctolitic magmas are excellent indicators of crustal contamination and have even be used to map out terrane boundaries at depth (Laramie - Mitchell et al., 1995).
The extended magmatic histories of several PAC (Nain-50 myr; Korosten-30 myr) imply the presence of long-lasting thermal anomalies in the mantle. Combined melting of the upper mantle/lower crust initiated by mantle plumes and/or extension following collision and crustal thickening, coupled with polybaric fractionation of plagioclase-rich magmas and extensive contamination during ascent is a likely mechanism for the formation of most anorthosites. In the case of Nain and Laramie, the presence of Paleoproterozoic suture zones, representing major lithospheric weaknesses, facilitated the ascent of plagioclase-rich diapirs and troctolitic magmas. Structural relief on the crust/mantle interface in the region of the suture zone may play an important role in the generation and ponding of magma during the earlier stages of magmatism. High resolution geophysical surveys across terrane boundaries characterized by Proterozoic anorthosites should be considered as a major research tool in the future.
Scoates JS & Chamberlain KR, J. Geol., 105, 331-343, (1997).
Hamilton MA, Ryan AB, Emslie RF & Ermanovics IF, Geol. Surv. Canada Curr. Res., 1998-F, 23-40, (1998).
Mitchell JN, Scoates JS & Frost CD, Contrib. Mineral. Petrol., 119, 166-180, (1995).
3218
Combined mapping, petrological, geochemical, isotopic and precise U-Pb geochronological (zircon/baddeleyite) studies within the unmetamorphosed Nain Plutonic Suite (NPS), Labrador, and the Laramie Anorthosite Complex (LAC), Wyoming, has resulted in significant advances in our understanding of the origin of massif anorthosite, related troctolitic and ferrodioritic rocks, and spatially-associated monzonitic-granitic suites. Both complexes are Mesoproterozoic in age and were emplaced across major Paleoproterozoic terrane boundaries that suture earlier Paleoproterozoic and Archean gneisses.
In Labrador, a working paradigm has been established (Emslie et al., 1994) wherein mantle-derived basaltic magmas, ponded and underplated at the base of the crust, initiated lower crustal melting (extracted as 1.351-1.287 Ga high Fe/Mg granitoids). The resulting hot, dry, plagioclase-rich granulitic residues were readily assimilated by the basaltic magmas, which helped maintain plagioclase on the liquidus for protracted periods. The large volumes of anorthositic magmas in the NPS emplaced between 1.330-1.295 Ga are therefore expected to be far from equilibrium with upper mantle compositions. Recent wide-angle seismic studies offshore the NPS (Reid, 1998) reveal a limited amount of crustal extension, and importantly, a high-velocity layer (~7.4 km/s) at the base of the crust which may confirm a fossil region of magmatic underplating directly related to NPS magmatism.
Nd-Sr-Pb isotopic studies indicate substantial mantle-crust interaction for both massif anorthosite complexes, resolvable because, relative to Grenville-aged massifs, they are emplaced into significantly older, isotopically-contrasted crust. Troctolitic intrusives typically represent the least-contaminated magmatic members, though rarely do they represent purely isotopically-unaffected derivatives of a depleted mantle source. However, suites of c. 1.280 Ga diabase dykes, immediately succeeding NPS plutonism, have initial <epsilon> Nd values up to +4.2 and imply that much of northern Labrador was probably underlain by isotopically depleted mantle at the time.
Paradoxes still remain: 1) Maximum magma production rates calculated for NPS basic rocks, allowing for unseen cumulates in the mid-crust and a seismically-constrained underplated region, are resultingly very low in comparison with most continental flood basalts; 2) LAC and NPS were emplaced in regions which had experienced remarkably similar but lesser volumes of anorthositic+granitic±dioritic magmatism during the Paleoproterozoic; 1.76 Ga Horse Creek anorthosite complex predates LAC by 330 Myr (Scoates and Chamberlain, 1997), while an important 2.14-2.11 Ga complex of nearly identical character predates NPS magmatism by 800 Myr (Hamilton et al., 1998); 3) Models where anorthositic magmatism is a consequence of asthenospheric melts that replace tectonically-thickened lithospheric roots removed through detachment, convection or plume erosion have time-scales of 30-400 Myr - what are the controlling factors?
Emslie RF, Hamilton MA & Theriault RJ, J. Geol, 102, 539-558, (1994).
Reid I, ECSOOT/Lithoprobe, (in press).
Scoates JS & Chamberlain KR, J. Geol, 105, 331-343, (1997).
Hamilton MA, Ryan AB, Emslie RF & Ermanovics IF, Geol. Surv. Canada Curr Res, 1998-F, 23-40, (1998).
Mapping and Ni-Cu exploration of anorthosite and coeval mafic intrusions in eastern Laurentia are currently based on classical models of diapir, layered intrusion and dyke emplacement in anorogenic settings. Field-based research and deep seismic profiles in western Grenville reveal that the Morin anorthosite suite and the coeval Chevreuil monzonite-gabbro suite were emplaced in a compressive regime during renewed orogenesis after early 1.2 Ga terrane accretion to Proto-Laurentia. The 1.17-1.16 Ga Chevreuil suite consists of porphyritic monzonite plutons (<55 km2), vertically-layered gabbroic stocks (5-30 km2), gabbro sheets (~10 km2), mixed felsic/mafic monzonite-diorite sheet-like intrusions (1-55 km2) and a swarm of microdiorite dykes. Eastward in the Morin terrane, coeval porphyroid monzonite marked the onset of the Morin AMCG magmatism. Previously considered syn-metamorphic, the Morin and Chevreuil suites posdate migmatization. The Chevreuil bodies were emplaced in zones of crustal weakness across mechanically contrasting felsic gneiss, marble and quartzite supracrustal domains. Arrays of sheet-like monzonite-gabbro intrusions define magmatic corridors along and concordant with the western, northern and eastern tectonic boundaries of the host supracrustal belt and along the high-strain zones between its lithotectonic domains. The mafic-felsic sheets and their concordant dyke swarms record repeated magma channeling and mingling, locally under high magma pressure. Under the Morin anorthosite east-dipping, discontinuous reflectors reach a highly reflective crust near the Moho. The reflectors are interpreted to be pre-Chevreuil extensional collapse structures which tapped underplated felsic and mafic magmas reservoirs. The 30 Ma hiatus between peak metamorphism and onset of Morin AMCG and Chevreuil magmatism is consistent with a post-collisional intraplate extensional magmatic event. Kinematic indicators however, record regional-scale E-W flattening at the late-stage of the Chevreuil magmatism. Regional compression is favorable for magma emplacement, the syn-tectonic magmas being emplaced at high angle to the regional shortening direction in the presence of strong anisotropy and high magma pressure (Lucas and St-Onge 1995). Considering the short hiatus for the Chevreuil magmatism and the connection between magma emplacement and major crustal structures, we advocate that compression was the effective regime throughout the 1.17-1.15 Ga time span, and consequently for the onset of Morin AMCG magmatism. We advocate that the Chevreuil suite is a case-example of post-collisional composite gabbroic and monzonitic magmatism representing a lateral off-shoot of AMCG magmatism. Although slab delamination and asthenosphere rise likely instigated melting, leading to build up of voluminous and extensive magma reservoirs, magmas likely stagnated at the crust-mantle boundary until ascent was triggered by renewed orogenic activity. Crustal anisotropy such as tectonic boundaries would provide preferential pathways for magmas, hence the connection of gabbro-monzonite and AMCG petrogenesis with major crustal structures. These suites are the expression of post-collisional melt/magma transfer from source to surface during renewed orogenesis.
Lucas, S and St-Onge, M, J. Struct. Geol., 17, 475-491, (1995).
2855
Zircon is the principal mineral in the dating of granites. One of the limitations is related to the high melting point of the mineral. This entails the possibility that the U-Pb systems of xenocryst zircons remain undisturbed despite enclosure in granitic melts. Zircons of xenocrystic types are quite common, for instance, in rapakivi granites (se e. g. Claesson et al., 1997). To assess the influence of the physical and chemical parameters of granitic melts on the U and Pb isotopes in xenolithic zircons, a set of experiments in a zircon-granitic melt system was conducted. The employed zircons belonged to an igneous -type population with a concordant U-Pd age of 2040±5 Ma and a structure comprising both metamict and non-metamict parts.
Two series of experiments were set up, both at P=3 kbar but at temperatures of 1250 and 850°C, respectively. A high-pressure apparatus of a gas-bomb-type with internal heating was used. The durations of the individual experiments in each of the two series were 8, 12, 17, 32, and 72 hours. After the experiments, U-Pb isotope analysis was carried out, while x-ray, infrared spectroscopy and thermoluminescence were employed to elucidate the particulars in changes in the zircon structures and relate this to the degree of disturbance of the U-Pb isotope systems (Bibikova et al., 1998). The conclusions are:
- The behaviour of U-Pb isotope systems in the zircons exposed to granitic melt was defined by the two-phase, metamict-non metamict structure of the zircon crystals and, particularly, by different stability of radiogenic lead in these two parts of the structure.
- Recrystallization of the metamict parts of the zircons occurred even in the shortest-duration experiments, which was confirmed by all the crystallochemical approaches employed. The recrystallization was accompanied by significant loss of lead.
- In contrast, the loss of lead from the crystalline, non-metamict parts of the zircon crystals was negligible, being governed by the law of volume diffusion in crystalline matter.
- The preservation of radiogenic Pb in the parts of the zircons marked by crystalline, non-metamict structure even during a long residence time in granitic melt explains the preservation of isotope memory in zircon xenocrysts in granites in general and in rapakivi-granites in particular.
This work was supported by the RFBR, project No. 97-05-65565a.
Claesson S., Anderson U., Schuhmacher M., et al, Terra nova, Abst. Suppl., 9:1, 356, (1997).
Bibikova E, Bykov I, Ivliev A et al, Geochemistry Inter, 36, 40-46, (1998).
Anorthosite-Rapakivi Granite Petrogenesis in Connection with Major Crustal Structures
Anorthosite-Rapakivi Granite Petrogenesis in Connection with Major Crustal Structures
Anorthosite-Rapakivi Granite Petrogenesis in Connection with Major Crustal Structures
Anorthosite-Rapakivi Granite Petrogenesis in Connection with Major Crustal Structures
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications