Journal of Petrology Advance Access originally published online on December 10, 2004
Journal of Petrology 2005 46(4):671-699; doi:10.1093/petrology/egh093
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pan-African Tectonism in the Western Maud Belt: P–T–t Path for High-grade Gneisses in the H.U. Sverdrupfjella, East Antarctica
1 DEPARTMENT OF GEOLOGICAL SCIENCES, UNIVERSITY OF CAPE TOWN, RONDEBOSCH 7701, SOUTH AFRICA
2 RESEARCH SCHOOL OF EARTH SCIENCES, AUSTRALIAN NATIONAL UNIVERSITY, CANBERRA, A.C.T. 002, AUSTRALIA
RECEIVED MARCH 3, 2004; ACCEPTED OCTOBER 22, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Extensive high-grade polydeformed metamorphic provinces surrounding Archaean cratonic nuclei in the East Antarctic Shield record two tectono-thermal episodes in late Mesoproterozoic and late Neoproterozoic–Cambrian times. In Western Dronning Maud Land, the high-grade Mesoproterozoic Maud Belt is juxtaposed against the Archaean Grunehogna Province and has traditionally been interpreted as a Grenvillian mobile belt that was thermally overprinted during the Early Palaeozoic. Integration of new U–Pb sensitive high-resolution ion microprobe and conventional single zircon and monazite age data, and Ar–Ar data on hornblende and biotite, with thermobarometric calculations on rocks from the H.U. Sverdrupfjella, northern Maud Belt, resulted in a more complex P–T–t evolution than previously assumed. A c. 540 Ma monazite, hosted by an upper ampibolite-facies mineral assemblage defining a regionally dominant top-to-NW shear fabric, provides strong evidence for the penetrative deformation in the area being of Pan-African age and not of Grenvillian age as previously reported. Relics of an eclogite-facies garnet–omphacite assemblage within strain-protected mafic boudins indicate that the peak metamorphic conditions recorded by most rocks in the area (T = 687–758°C, P = 9·4–11·3 kbar) were attained subsequent to decompression from P > 12·9 kbar. By analogy with limited U–Pb single zircon age data and on circumstantial textural grounds, this earlier eclogite-facies metamorphism is ascribed to subduction and accretion around 565 Ma. Post-peak metamorphic K-metasomatism under amphibolite-facies conditions is ascribed to the intrusion of post-orogenic granite at c. 480 Ma. The recognition of extensive Pan-African tectonism in the Maud Belt casts doubts on previous Rodinia reconstructions, in which this belt takes a pivotal position between East Antarctica, the Kalahari Craton and Laurentia. Evidence of late Mesoproterozoic high-grade metamorphism during the formation of the Maud Belt exists in the form of c. 1035 Ma zircon overgrowths that are probably related to relics of granulite-facies metamorphism recorded from other parts of the Maud Belt. The polymetamorphic rocks are largely derived from a c. 1140 Ma volcanic arc and 1072 ± 10 Ma granite.
KEY WORDS: Maud Belt; Pan-African orogeny; geochronology; P–T–t path, East Antarctica
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The East Antarctic Shield consists of a series of Archaean nuclei surrounded by polydeformed Proterozoic metamorphic terrains of complex tectono-thermal history (Krynauw, 1996
; Fitzsimons, 2000b). Radiometric age data clustering between 1300 and 900 Ma (traditionally compared with the ‘Grenvillian’ of North America) and between 600 and 450 Ma (the ‘Pan-African’ of Africa) have been reported from these polymetamorphic terrains throughout East Antarctica (Fitzsimons, 2000b, and references therein). The widespread occurrence of the former age group was interpreted as evidence of a continuous Grenville-age mobile belt. This so-called Circum East Antarctic Mobile Belt was linked with the Grenville Belt of North America and the resulting global-scale orogen considered the main suture in Rodinia reconstructions (Hoffman, 1991; Moores, 1991). Work in the Lützow–Holm Bay area (Shiraishi et al., 1994), the Larsemann Hills (Dirks et al., 1993), the Prydz Bay area (Fitzsimons et al., 1997) and in parts of central Dronning Maud Land (Jacobs et al., 1998, 2003), however, has led to the recognition that Pan-African tectonic overprinting in East Antarctica is of a greater intensity and importance than was previously thought, thus challenging the hypothesis of a single continuous Grenville-age mobile belt around East Antarctica (Fitzsimons, 2000a).
Two main Pan-African orogenic belt systems have been recognized in East Antarctica (Fig. 1). The Prydz Belt is considered to extend from the Leeuwin Complex in Western Australia to Prydz Bay and then inland (Boger et al., 2001), whereas the Lützow–Holm Belt (Fitzsimons, 2000b) is regarded as an extension of the East African Orogen into East Antarctica, where it passes through Lützow–Holm Bay and central Dronning Maud Land (Shiraishi et al., 1994; Grunow et al., 1996; Shackleton, 1996; Jacobs et al., 1998). The role of the high-grade metamorphic rocks that make up the Maud Belt of western Dronning Maud Land in the tectonic evolution of East Antarctica remains elusive.
|
Both a thermal and a tectono-thermal model have been proposed for the Pan-African history of the Maud Belt. In the former, the belt is considered a Grenville-age orogenic belt that experienced purely thermal overprinting during Pan-African times (Groenewald et al., 1995
), whereas, in the latter, considerable orogenic deformation is assumed to have taken place during Pan-African times (Jacobs et al., 1998, 1999, 2003). Clarification of this issue is pivotal for any attempt to reconstruct the palaeogeography during assembly and break-up of an inferred Mesoproterozoic supercontinent (be it Rodinia or any other configuration) and subsequent amalgamation of Gondwana.
Evidence of a purely thermal, Pan-African overprint stems largely from the central to northeastern Maud Belt, i.e. the nunataks of Kirwanveggen and H.U. Sverdrupfjella (Fig. 2), where the penetrative fabric, resulting from top-to-NW thrusting, has been interpreted as Grenvillian in age (Grantham et al., 1995; Groenewald et al., 1995). In contrast, strong evidence for Pan-African tectonic activity has been provided for the southwestern part of the belt by Golynsky & Jacobs (2001), who argued on structural and aeromagnetic grounds for a major shear zone there (Heimefront Shear Zone, Fig. 2) to be the continuation of the East African orogenic front.
|
The aim of this study is to examine the age of the penetrative fabric in the northern portion of the Maud Belt, specifically in southern parts of the eastern H.U. Sverdrupfjella, where the best exposed and most complete cross-section through the tectonostratigraphic units of the northern Maud Belt can be found. To that effect, new sensitive high-resolution ion microprobe (SHRIMP) and conventional U–Pb single zircon and monazite ages as well as Ar–Ar ages on hornblende and biotite are presented, together with detailed petrography and geothermobarometry, to generate a pressure–temperature–time (P–T–t) path for the high-grade gneisses in that area.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Western Dronning Maud Land extends from
70°S to 75°S and from 15°W to 03°E. It consists of two geological provinces: the Archaean to Mesoproterozoic Grunehogna Province and the Mesoproterozoic to early Palaeozoic Maud Belt (Krynauw, 1996). The Grunehogna Province, which is interpreted as a portion of the Kalahari Craton that became detached during Gondwana break-up, is assumed to be in tectonic juxtaposition with the Maud Belt along an inferred crustal discontinuity that is masked by the Pencksökket and lower Jutulstraumen glaciers (Fig. 2).
In the mountain range and isolated nunataks of H.U. Sverdrupfjella (Fig. 3), the structurally lowest exposed unit consists of intercalated tonalitic orthogneiss and hornblende–biotite gneiss (Jutulrøra Formation). It is interpreted as a volcanic arc sequence of interlayered calc-alkaline volcanic and clastic sedimentary rocks on the basis of geochemical and field characteristics (Groenewald et al., 1995). This is followed by a sequence of compositionally banded, in places migmatitic, biotite–hornblende gneiss with highly variable proportion of felsic and mafic minerals and dominance of plagioclase over K-feldspar, and intercalated discontinuous, boudinaged lenses of calcitic marble, calc-silicate, and amphibolite, all of which constitutes the Fuglefjellet Formation (Fig. 4).
|
|
Semipelitic garnet–biotite gneiss, quartzofeldspathic paragneiss and intermediate to felsic orthogneiss with subordinate amphibolite and calc-silicate enclaves (Rootshorga Formation) dominate the outcrops along the entire length of the eastern H.U. Sverdrupfjella. The banded nature and conformity with the layering in the paragneiss points to a volcanic rather than plutonic origin of the orthogneiss units. Although previous workers (Grantham et al., 1995
; Groenewald et al., 1995) inferred a tectonic contact between the Fuglefjellet and Rootshorga Formations, fieldwork by the authors revealed a gradational transition from the former to the latter formation in a low-strain zone NE of Skarsnuten (Fig. 3). This transition is characterized by an increase in the biotite/hornblende ratio within the paragneiss and a decrease in the amount of carbonate and calc-silicate intercalations.
A retroarc basin is envisaged for the depositional environment of the Fuglefjellet and Rootshorga Formations on the basis of their respective field relationships and geochemistry (Groenewald et al., 1995; Board, 2002). A Rb–Sr whole-rock age of 1145 ± 49 Ma for a metavolcanic unit within the banded gneisses of the Jutulrøra Formation (A. Moyes, unpublished data, 1986), a U–Pb SHRIMP zircon age of 1131 ± 25 Ma (Harris et al., 1995) for a sample of pre-tectonic garnet–biotite granitic gneiss of calc-alkaline composition, and a U–Pb single zircon age of 1131 ± 7 Ma for a related felsic tuff layer in the corresponding foreland deposits of the Ritscherflya Supergroup (Frimmel, 2004) constrain the age of the volcanic island arc.
Slightly younger (1103 ± 17 Ma; Harris et al., 1995) deformed tabular, in places distinctly megacrystic, granite gneiss bodies (Sveabreen Suite) display a within-plate geochemical affinity (Groenewald et al., 1995; Board, 2002). Several generations of deformed mafic intrusions are distinguished. The majority of the mafic bodies occur as lenses, boudins, concordant layers (sills or lava flows) and rarely dykes. They pre-date the deformation and are intrusive into the oldest pre-tectonic granite.
Post-tectonic plutons, sheets and dykes of monzogranitic composition (Brattskarvet Suite) are widespread. Mesozoic alkaline complexes (170 ± 4 Ma) and undeformed Jurassic dolerite dykes (Grantham et al., 1988; Harris & Grantham, 1993) are related to the break-up of Gondwana and reflect the youngest magmatic pulse in the area.
Five phases of deformation (D1–D5) have been recognized (Grantham et al., 1995). Transposition of earlier structures is evident in the presence of numerous rootless intrafolial folds, ascribed to D1. The penetrative fabric (S2) is shallow east-dipping (in the western domain) to moderately SE-dipping (in the eastern domain) and related to east- to SE-plunging tight to isoclinal folds and sheath folds, a moderately east- to SE-plunging mineral elongation lineation, and dominant top-to-NW shear sense (D2). The orientation of the majority of fold axes corresponds very closely to that of the mineral elongation lineations and the fold axes determined from foliation data. Regional warping and local NE–SW-trending open folds, in places leading to a top-to-SE reverse shear sense, reoriented the S2 fabric slightly and led to localized fabric transposition defined by biotite (D3). Localized large-scale open, upright folding on NE-trending axes reflect D4 in the western and northeastern parts of the H.U. Sverdrupfjella, whereas in the southern domain, regional-scale gentle warping of the SE-dipping S2 fabric and D2 high-strain zones about moderately SE-plunging axes mark D4. The development of steep to vertical normal faults and joints with north–south and NE–SW trends is the youngest deformation recognized in the H.U. Sverdrupfjella (D5).
Both D1 and D2 have been assigned a Grenvillian age by Grantham et al. (1995) based on the assumption that the Sveabreen granitic gneiss was emplaced post-D1 and syntectonically with respect to D2. Localized fabric development within the Brattskarvet Suite has led some of those workers to interpret D3 and D4 as Pan-African in age. However, a magmatic origin of the rather chaotic and multi-directional structures in the Brattkarvet pluton is considered more likely. The similarity in the orientation of the stress fields for the D1 to D4 deformations makes a clear distinction difficult in many places and their age remains uncertain. There is general agreement that the youngest deformation, D5, is related to the Mesozoic break-up of Gondwana (Groenewald et al., 1995).
The metamorphic history of the area is complex. Evidence of the earliest stage of metamorphism (M1) comes from cores of pre-tectonic mafic boudins that display a granulite-facies mineral assemblage for which P–T conditions of 8–10 kbar and 850°C have been estimated (Groenewald & Hunter, 1991). The majority of the rock types throughout the H.U. Sverdrupfjella consist of syn-kinematic (syn-D2) mineral assemblages characteristic of the upper amphibolite facies (Fig. 4). Metapelitic gneisses show abundant evidence of partial melting with pre-, syn- and post-tectonic leucosome domains. In contrast, no evidence of partial melting exists within the mafic rocks. In the eastern area, Grantham et al. (1995) distinguished an early high-pressure stage (P = 9–20 kbar; T = 720–850°C) from a subsequent near-isothermal decompression to conditions transitional between amphibolite and granulite (P = 5–8 kbar; T = 600–790°C), retrograde rehydration at similar P–T conditions, all of which constitute M2, and localized late-stage greenschist-facies conditions (M3). Evidence of the latter is confined to syn-D5 fractures and joints and the wall rock of undeformed dolerite dykes.
In the western area, the distinction between the various metamorphic events is problematic, as only amphibolite-facies conditions (P = 3–9 kbar; T = 620–700°C) have been described (Grantham et al., 1995). Some of the above P–T constraints are rather loose and problematic, because of the use of unreliable geothermobarometers and potential disequilibrium among the mineral assemblages used, but a recent re-examination of the metamorphic conditions recorded by amphibolites there indicated peak temperatures around 730°C (Grosch & Frimmel, 2004). Grantham et al. (1995) interpreted the apparent differences in structural style and metamorphic evolution between the western and eastern H.U. Sverdrupfjella as indicating that the rocks of the latter were thrust over those of the former during D2, but the more recent results (Grosch & Frimmel, 2004) do not support this.
|
PETROGRAPHY AND MINERAL CHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Mineral assemblages in the felsic gneiss, semi-pelitic gneiss, carbonate and calc-silicate units (Fig. 4) provide only limited constraints on the metamorphic history. However, mineral assemblages and disequilibrium textures developed in the cores of competent anhydrous metabasic rocks, which form resistant tectonic lenses and boudins, and in the metapelitic rocks are more useful to place constraints on the P–T evolution of the area. Six amphibolite samples, representing the range of mafic rock types in the southern H.U. Sverdrupfjella, and three samples of metapelitic gneiss were therefore chosen for mineral chemical analysis. The mineral chemical data (Electronic Appendix 1, which may be downloaded from the Journal of Petrology website at http://www.petrology.oupjournals.org) were acquired by conventional electron microprobe analysis using a Cameca Camebax instrument at the Department of Geological Sciences, University of Cape Town. The instrument was run at an accelerating voltage of 15 kV [25 kV for the analysis of rare earth elements (REE) in monazite] and a beam current of
40 nA. Counting times at peak and background positions were 10 s, with the beam defocused to 2–3 µm. Natural and synthetic standards were used for calibration and the Cameca PAP correction procedure was employed. The detection limits for the elements analysed for are between 0·02 and 0·09%. Mineral abbreviations used are from Kretz (1983). All Fe is expressed as Fe2+ unless otherwise stated.
Metabasic rocks
The oldest mineral assemblage recorded within the cores of some of the pre-tectonic mafic rocks is Grt–Cpx–Qtz–Rt. Little of the original assemblage is preserved, however, because of the extensive re-equilibration of these rocks during subsequent metamorphism that led to the widespread development of the assemblage Hbl–Pl–Ilm–Spn. In most of the pre-tectonic mafic rocks, the presence of garnet and clinopyroxene is mainly inferred from the occurrence of pseudomorphs of plagioclase and hornblende. However, earlier garnet and, very rarely, clinopyroxene are still preserved in cores of resistant tectonic lenses. That garnet occurs as sub-rounded to irregularly shaped grains (1–4 mm in size), contains inclusions of rutile and quartz, and is commonly surrounded by a symplectitic corona of hornblende and plagioclase (Fig. 5b). Clinopyroxene occurs as irregularly shaped grains (>3 mm in length) that are generally separated from the garnet and the matrix by a rim of either hornblende or hornblende–plagioclase symplectite. Plagioclase exsolution lamellae constitute up to 45 vol. % of the clinopyroxene grains, and small patches of hornblende occur in places, formed at the expense of clinopyroxene. In one sample (WBSV079), however, granoblastic clinopyroxene relics that are free of plagioclase exsolution lamellae appear to be in textural equilibrium with garnet, hornblende and plagioclase (Fig. 5a).
|
Most amphibolite samples are dominated by hornblende and plagioclase. These minerals typically form symplectitic intergrowths where they surround relics of the older Grt–Cpx assemblage (Fig. 5b), and have a granoblastic texture where they occur as matrix minerals. The presence of symplectitic intergrowths and clinopyroxene grains with plagioclase exsolution lamellae suggests limited fluid infiltration into the core of those metabasic rocks in which such textures are preserved. Rutile and ilmenite, in places rimmed by titanite, occur both within the matrix and as inclusions within hornblende. Titanite is in textural equilibrium with the hornblende and plagioclase, which define a fabric in the foliated outer portions of the pre-tectonic metabasic boudins that corresponds to the regional S2.
The above textures suggest decompression from eclogite-facies to upper amphibolite-facies conditions by the continuous reaction
 | (1) |
The large proportion of plagioclase exsolution lamellae in the clinopyroxene is a typical feature of decompressed eclogites (Boland & van Roermund, 1983) and indicates a precursor of omphacitic composition that expelled jadeite and tschermakite components by the reaction
 | (2) |
The absence of a penetrative fabric associated with the Grt–Cpx assemblage suggests a pre-tectonic (pre-D2) origin for this assemblage, although this could simply be due to the competent nature of the mafic bodies, making such rocks generally more resistant to deformation than the host paragneiss. The good preservation of the decompression textures described above points to the eclogite- and amphibolite-facies assemblages having formed during the same tectonic episode, whereby metamorphic recrystallization outlasted deformation (D2).
Widely spaced biotite laths cut across the above mineral assemblages. The abundance of biotite laths generally increases from the cores (<<1 vol. %) to the edges (up to 20 vol. %) of the mafic bodies, although the cores of many of the more re-equilibrated rocks contain up to 15 vol. % biotite. This distribution is consistent with biotite formation being related to potassic fluid infiltration during a retrograde metamorphic stage. Although some of the biotite grains are aligned parallel to S2, most are randomly oriented. Both textural types of biotite have the same composition. The metabasic rocks do not display an S3 fabric, and the latest metamorphic imprint recognized is the local replacement of biotite by chlorite near syn-D5 fractures and joints.
A total of 476 electron microprobe analyses were obtained of garnet, clinopyroxene, plagioclase, ilmenite, amphibole and biotite. Representative mineral analyses for each sample are given in Tables 1 and 2. The complete dataset may be downloaded from the Journal of Petrology website at http://www.petrology.oupjournals.org/ (Electronic Appendix 1).
|
|
Garnet is present (10–55 vol. %) in all of the metabasic samples analysed and consists predominantly of almandine (XAlm = 46–62%, XGro = 17–31%, XPyr = 11–23%, XSps = 0·6–4·3%). XFe = Fe/(Fe + Mg) values range from 0·67 to 0·84. Garnet is mostly homogeneous. In some grains, XSps displays a slight decrease with a corresponding increase in XPyr from core to rim, with XGro remaining constant.
Clinopyroxene (up to 30 vol. %) is homogeneous and contains between 19·4 and 24·1 wt % CaO, with 2–7% jadeite, up to 9% calcium-tschermakite and up to 1% aegirine contents. XFe varies considerably between samples (0·26–0·53) but is relatively uniform within samples.
Plagioclase is present in variable proportions (5–35 vol. %) in all of the analysed samples. Individual grains are homogeneous, but up to three compositionally different varieties are present within a given sample. Their anorthite contents are An60–71, An39–54 and An27–33 [where An = Ca/(Ca + Na)], with the more sodic compositions present in exsolution lamellae in clinopyroxene and in matrix plagioclase within well-equilibrated, symplectite-free samples. The orthoclase proportion is generally <3%.
Ilmenite is unzoned and occurs in minor quantities (<5 vol. %). The Fe/Ti ratio ranges from 0·906 to 1·004 and there is <0·05 Mg per formula unit FeTiO3.
Amphibole is a major constituent (10–40 vol. %) of all analysed samples. It shows little intra-sample compositional variability, but significant differences between samples, with Fe/(Fe + Mg) ranging from 0·42 to 0·72 (Table 2). Assuming all iron as Fe2+, the amphibole grains range from pargasite to ferrohornblende and ferropargasite.
Biotite is present in most samples analysed (<10 vol. %). Biotite laths that are aligned parallel to S2 contain between 2·0 and 5·6 wt % TiO2 and up to 0·98 wt % F (Table 2). As with the other ferromagnesian minerals, biotite shows considerable inter-sample compositional variation, with Fe/(Fe + Mg) ranging from 0·37 to 0·72. Randomly oriented biotite is compositionally indistinguishable from the syn-tectonic biotite.
Metapelitic gneiss
The peak (and earliest recognizable) metamorphic assemblage in the metapelitic rocks is Grt–Bt–Ky–Pl–Kfs–Qtz–Gr, corresponding to the upper amphibolite facies. Aligned elongate laths of kyanite, biotite and graphite define the S2 fabric and are set in a matrix of strained quartz, plagioclase and K-feldspar. Garnet occurs as porphyroblasts, 2–5 mm in diameter, that are wrapped by the foliation. In high-strain zones the garnet occurs as 
-porphyroclasts that indicate a top-to-NW shear sense. Quartz, rutile, biotite, monazite, and zircon occur as inclusions in the garnet. The internal fabric is continuous with the external S2, from which syn-tectonic (with respect to D2) garnet growth is inferred. Kyanite grains vary in size from 0·5 mm to 5 mm, exhibit localized kinking and associated undulose extinction, and contain inclusions of biotite, quartz, monazite, zircon and rutile. No evidence of an earlier metamorphic event is preserved in the metapelite samples.
Retrograde metamorphic effects include sillimanite growth at the expense of kyanite, and muscovite growth at the expense of kyanite and/or sillimanite. In sillimanite- and muscovite-bearing domains, garnet is partially replaced by intergrowths of biotite and plagioclase. Rare local replacement of biotite by chlorite is limited to areas adjacent to syn-D5 joints and fractures.
A total of 182 electron microprobe analyses were obtained of garnet, biotite and plagioclase (for representative analyses see Table 3). Garnet composition is dominated by almandine (XAlm = 72–79, XPyr = 12–20, XGro = 4–7, XSps = 2–4%). Its modal proportion is typically between 5 and 20 vol. %. In a particularly garnet-rich (45 vol. %) sample, a markedly higher grossular content (8–25%) was noted. Garnet grains in most samples are not zoned. The only exception is a systematic enrichment in XGro and corresponding depletion of XAlm from core to rim in the garnet-rich sample, which might be a relic of prograde growth zonation reflecting progressive compression. Plagioclase constitutes between 5 and 15 vol. %. It is unzoned, has An27–41 and XOr of <0·015.
|
Modal proportions of biotite range from 5 to 30 vol. %, only exceeded in thin intercalations of biotite–garnet layers (40 vol. %). Aligned and randomly oriented biotite grains are compositionally similar (XFe = 49–55%), both within and between samples (Table 3). Biotite in the biotite–garnet layers is more ferrous.
|
GEOTHERMOBAROMETRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Metabasic rocks
The application of geothermobarometry to the metabasic rocks is complicated by their multi-stage metamorphic history. Quantifying P–T for the early metamorphic stage is problematic, as the eclogite-facies assemblage was largely replaced by an amphibolite-facies assemblage, and the compositions of the original garnet and clinopyroxene most probably have changed.
Assuming that clinopyroxene containing albite-rich plagioclase exsolution lamellae represents the retrogressed equivalent of former omphacite (Boland & van Roermund, 1983) and that this replacement was isochemical, original omphacite compositions were approximated by reintegration of clinopyroxene and plagioclase volumes and compositions. Sample WBSV034 was found most suitable for this calculation, which follows that outlined by Markl & Bucher (1997). The modal abundances of plagioclase exsolution lamellae and host clinopyroxene were determined by point counting in texturally homogeneous parts of the section as respectively 43 and 57 vol. % clinopyroxene. The recalculated omphacite contains approximately 30% jadeite and 14–17% calcium tschermakite (Table 4). A minimum pressure of 12·9–14·3 kbar is estimated from these compositions by modelling the reaction
 | (3) |
).
|
Textural evidence of newly formed hornblende and plagioclase (of intermediate composition) being in equilibrium with the partially resorbed garnet in several of the metabasic rock samples opens up the possibility of placing constraints on the amphibolite-facies P–T conditions. Textural disequilibrium between biotite and the earlier-formed minerals precludes the use of geothermobarometry in the quantification of the retrograde P–T conditions.
The Grt–Hbl and Hbl–Pl geothermometers of Graham & Powell (1984) and Holland & Blundy (1994), and the Grt–Hbl–Pl–Qtz and Grt–Rt–Ilm–Pl–Qtz geobarometers of Kohn & Spear (1991) and Bohlen & Liotta (1986) were used. The edenite–tremolite geothermometer of Holland & Blundy (1994) was applied to all of the samples except WBSV066, for which the edenite–richterite geothermometer was preferred because of the absence of quartz. In each sample, P–T estimates were made for several microdomains in which the relevant phases appear to be in textural equilibrium. Samples WBSV034 and WBSV079 both contain clinopyroxene but only in the latter does this mineral appear to be in equilibrium with the amphibolite-facies assemblage. Consequently, the Grt–Cpx geothermometer (Ellis & Green, 1979; Krogh, 1988; Pattison & Newton, 1989; Ai, 1994) and Grt–Cpx–Pl–Qtz geobarometer (Newton & Perkins, 1982; Moecher et al., 1988; Eckert et al., 1991) were applied in an attempt to place further constraints on the P–T conditions. P–T estimates were obtained from several microdomains containing garnet and clinopyroxene, as well as such domains, in which garnet, plagioclase, clinopyroxene and quartz appear to be in textural equilibrium, using different combinations of the above geothermobarometers. Results of the application of the various geothermobarometers are summarized in Tables 5 and 6.
|
|
In less equilibrated, symplectite-rich samples (WBSV034 and WBSV041), temperature estimates determined by the Grt–Hbl geothermometer are at variance with those obtained from the Hbl–Pl geothermometers. In these samples Fe–Mg exchange equilibrium between hornblende and garnet may not have been attained and, therefore, results from the Hbl–Pl thermometer are favoured.
The temperature and pressure ranges determined for the various samples using the Hbl–Pl geothermometer in combination with the Grt–Hbl–Pl–Qtz and Grt–Rt–Ilm–Pl–Qtz geobarometers are within error, both within and between samples. Pressure estimates obtained for the various samples using the Fe and Mg end-members of the Kohn & Spear (1991) geobarometers display some variation that is explainable by slight differences in garnet and amphibole mineral chemistry.
The combination of the Pattison & Newton (1989) thermometer, regardless of which barometer was used, when applied to the Grt–Cpx–Pl–Qtz microdomains in sample WBSV079, consistently returned temperatures of 100–110°C lower than the estimates in Table 5. Pressure estimates obtained using the Newton & Perkins (1982) and Eckert et al. (1991) geobarometers are generally lower (between 1·5 and 3·0 kbar) than those determined using the various Grt–Hbl–Pl–Qtz–Ilm–Rt geothermobarometers (Table 5), regardless of which thermometer was used. Consequently, results from the Ellis & Green (1979) and Moecher et al. (1988) thermobarometers are favoured. The similarity between the P–T estimates determined using these calibrations and the estimates presented in Table 5 suggests that local equilibrium may have been attained between garnet, clinopyroxene, plagioclase and quartz during M2b in this sample. Combining the results of the various geothermobarometers results in an overall range in P–T estimates for M2b of T = 687–758°C and P = 9·4–11·3 kbar.
The above results were compared with those obtained by applying the average P–T method (Powell & Holland, 1994) to garnet-bearing amphibolite samples using THERMOCALC version 2.75 (Powell et al., 1998). Mineral analyses from microdomains in which all of the M2b phases appear to be in textural equilibrium were selected for optimal geothermobarometry, and the activities of the various end-members were calculated using the program AX of T. Holland (available at http://www.esc.cam.ac.uk/astaff/holland/index.html). Average P–T values were calculated using the activities determined for the various end-members of garnet, ilmenite, hornblende, clinopyroxene and plagioclase, as well as quartz, rutile and titanite, where present, at estimated metamorphic temperatures. Average P–T calculations for all samples except WBSV041 yielded -fit values slightly larger than one, yet less than a maximum value required to pass a 
2 test, indicating that the input data are broadly consistent with the final P–T estimate (Powell & Holland, 1994). The ferroactinolite end-member in sample WBSV041 was flagged as being a potential problem, because it has a high value of e*fact. Deletion of this end-member from the data file produced a set of results that passed the 2 test, and this revised P–T estimate was used.
Ignoring the effect of fluid pressure and composition resulted in average P–T estimates that are geologically unreasonable (in the range 800–1000°C and 11–13 kbar), with the associated errors being very large (up to ±100°C and ±3 kbar). Limited fluid activity is evident from the symplectitic textures (e.g. Messiga et al., 1990). Powell & Holland (1994) suggested that the average fluid composition probably corresponds to that for which a minimum -fit is obtained in P–T–X calculations. Consequently, average P–T estimates were modelled using different fluid compositions until a minimum value of -fit was obtained, resulting in XCO2 = 0·1 and XH2O = 0·2–0·4 for samples WBSV031, WBSV034 and WBSV079, and XCO2 = XH2O = 0·03 for WBSV041 (Table 7). Combining the results of the application of optimal geothermobarometry to the various samples results in an overall range in P–T estimates of T = 642–792°C and P = 7·7–11·9 kbar.
|
Metapelitic gneiss
The Grt–Bt thermometer (Holdaway, 2000
) and Grt–Ky–Pl–Qtz barometer (Holdaway, 2001) were applied to the selected metapelitic gneiss samples. Several microdomains in which garnet, biotite, kyanite (present only in samples WBSV025 and WBSV084), plagioclase and quartz appear to be in textural equilibrium were chosen for geothermobarometry. The employed geothermometer and geobarometer use the ferric iron estimates determined for natural reduced garnet [Fe3+/(Fe2+ + Fe3+) = 0·03] and biotite [Fe3+/(Fe2+ + Fe3+) = 0·116] from Maine (Guidotti & Dyar, 1991; Holdaway et al., 1997) and these estimates were also used in this study to maintain internal consistency. Thermobarometric calculations were made with the programs GB, GASP and GBGASP of Holdaway (2000, 2001), using T = 680°C and P = 7 kbar as input P–T values for samples WBSV025 and WBSV084, and T = 720°C and P = 7 kbar for sample WBSV110. The P–T data and ranges were calculated in a similar way to those obtained for the metabasic rocks. The results of the application of these geothermobarometers to the three samples of metapelitic gneiss are presented in Table 8. Estimates of amphibolite-facies P–T conditions in samples WBSV025 and WBSV084 are T = 677–693°C; P = 7·0–8·1 kbar, whereas T estimates for sample WBSV110, using various garnet rim and core compositions and assuming P = 7·0 and 8·0 kbar, range between 650–718°C and 725–767°C, respectively.
|
No meaningful P–T estimates could be obtained for a wide range of input conditions using the average P–T method. The lack of constraints on the fluid–melt phase hampers the application of the average P–T method to the metapelitic rocks, which have clearly been affected by fluid infiltration and partial melting.
P–T path
Three stages of metamorphic recrystallization are distinguished in the studied rocks. The first stage is evident only in garnet–clinopyroxene relics within strain-protected boudins that record a P > 12·9 kbar. No reliable temperature estimate could be derived from these relics, but from the widespread decompression textures preserved in many amphibolite samples, similar temperatures to those of the subsequent amphibolite-facies stage are inferred. Most of the mineral assemblages formed during the second stage, and P–T conditions calculated for individual metabasic rock samples, using various geothermobarometers, are all within the corresponding errors obtained from the application of the average P–T method (Table 7). The overall ranges in P–T estimates for the metabasic rocks are consistent with field observations and independent petrogenetic constraints. They plot between the water-saturated granite solidus and the vapour-absent muscovite melting reaction in the NKASH system, which controls the solidus relationships in the NCKFMASH system (White et al., 2001). This is in agreement with the field evidence of partial melting in the metapelitic rocks. The P–T estimates plot on the low-temperature side of the basalt solidus as determined for intermediate water activities (aH2O = 0·6; Holloway & Burnham, 1972). Low fluid activities derived for the pre-tectonic metabasic rocks are therefore consistent with a lack of partial melting in these rocks. Overall, a temperature between 687 and 758°C and a pressure between 9·4 and 11·3 kbar are considered the best estimates for this second metamorphic stage.
The, on average, slightly lower temperatures and pressures calculated for the metapelitic samples (Table 8) relative to the metabasic rocks indicate that the less competent metapelitic rocks have been subject to more extensive re-equilibration on the retrograde path. This is supported by the limited development of sillimanite at the expense of kyanite. As the garnet–biotite thermometer and the garnet–aluminium silicate–quartz–plagioclase barometer probably behave differently during the retrograde history as a result of the differences in relative efficiencies of component exchange between phases (e.g. Fe–Mg exchange is relatively faster and more efficient than Ca, Si and Al exchange), the P–T estimates (7·0–7·5 kbar, c. 690°C) may not represent a true point on the P–T path, but rather a combination based on P and T attained at different times. This retrograde overprint, representing the third metamorphic stage, affected the rocks after the fabric-forming deformation, as evident from the abundant post-tectonic growth of biotite in highly variable proportions in most rock types, including the rims of amphibolite boudins.
|
GEOCHRONOLOGY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGPETROGRAPHY AND MINERAL...GEOTHERMOBAROMETRYGEOCHRONOLOGYDISCUSSION AND CONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
U–Pb SHRIMP dating of zircon and monazite
Mineral separation was carried out on selected samples using standard heavy liquid and magnetic separation techniques. Individual zircon and monazite grains, and in situ monazite grains were analysed using the SHRIMP I and II ion microprobes at the Research School of Earth Sciences, Australian National University (ANU), Canberra. Cathodoluminescence (CL) imaging of zircon grains and backscatter electron (BSE) imaging of monazite grains was conducted prior to isotope analysis. SHRIMP analytical techniques have been described by Compston et al. (1984)
, Ireland & Gibson (1998) and Williams (1998). Isotopic ratios and inter-element fractionation were monitored by continuous reference to zircon (SL13) and monazite (from Thompson Mine, Canada) standards, fragments of which were mounted with each sample. The 206Pb/238U ratio in the unknown zircon grains was normalized to 0·0928 for SL13 (equivalent to an age of 572 Ma). 206Pb/238U ratios were normalized to 0·3152 for the monazite standard (1766 Ma). For all samples the SHRIMP analyses (Electronic Appendix 2) consisted of six scans through the mass range. Where 207Pb/206Pb ages are quoted or calculated, common Pb corrections were made using the measured 204Pb/206Pb composition and the relevant model Pb composition from Cumming & Richards (1975). For analyses with very low Th/U ratios (generally 
0·01), correction for common Pb using measured 208Pb was preferred. U–Pb ages were calculated from the 206Pb/238U compositions, with the correction for common Pb being made using the measured 207Pb/206Pb and 206Pb/238U as described by Compston et al. (1984). All ages quoted in this study were calculated using the recommended decay constants of Steiger & Jäger (1977). Statistical analysis of the data was conducted using the program ISOPLOT/EX (Ludwig, 2000). The isotope data were supplemented by in situ analyses of the REE concentrations in selected zircon and monazite grains, using electron microprobe and laser ablation inductively coupled mass spectrometry (LA-ICP-MS) techniques at the Department of Geological Sciences, University of Cape Town. For the latter, a Perkin Elmer Elan 6000 instrument, coupled to a Cetac LSX-200 LA module with a 266 nm frequency-quadrupled Nd–YAG laser was employed. The NIST 610 and 612 glass standards were used for zircon analyses, whereas an in-house standard of homogeneous monazite was used for the calibration of the monazite analyses. Lower limits of detection are typically below 20 ppb for the REE, and typical precision and accuracy range from 1 to 10 relative %.
Oldest pre-tectonic granitic gneiss (WBSV065)
The oldest pre-tectonic granitic gneiss forms the northernmost outcrop of the Skarsnuten group of nunataks (Fig. 3). It is overlain by the quartzofeldspathic rocks of the Fuglefjellet Formation and has been intruded by numerous pre-tectonic mafic bodies. Leucosome development is extremely limited and the granitic gneiss exhibits a weaker foliation than the overlying quartzofeldspathic gneiss.
All zircon grains display well-developed compositional growth zonation under CL imaging (Fig. 6), which is probably a primary magmatic structure (Vavra, 1990). Zircon overgrowths are generally absent, but irregularly shaped areas, characterized by bright CL, that are discordant to the growth zonation are present in a few of the grains. Results and their interpretation are given in Table 9 and Fig. 7. The general absence of overgrowths, in combination with the observation that this body is less intensely foliated than rocks of the overlying paragneiss sequence, suggests that it may have been an effective strain resistor during tectono-metamorphic events subsequent to its formation at 1132 ± 16 Ma.
|
|
|
Metapelitic gneiss (WBSV025)
The metapelitic gneiss of the Rootshorga Formation is part of a paragneiss package that overlies the granitic gneiss of the Jutulrøra Formation and that is cut by the pre-tectonic tabular gneissic granite of the Sveabreen Suite. Dark brown cores that are surrounded by relatively clear zircon overgrowths are visible in a few of the zircon grains in transmitted light. CL imaging revealed that the majority of the grains are homogeneous and do not display a complex internal structure (Fig. 6h). The cores that are visible under transmitted light in a few of the grains are generally indistinguishable from the rims on the CL images, although a few grains contain small irregularly shaped areas of brighter CL that are interpreted as highly resorbed cores. These features suggest that the unstructured zircon was probably formed as a result of almost complete dissolution of older zircon grains during a subsequent (metamorphic) event. The fact that non-metamict zircon remains closed to Pb diffusion at most crustal temperatures (Mezger & Krogstad, 1997
; Cherniak & Watson, 2000), together with the observation that the irregularly shaped older cores, where present, are locally crosscut by younger zircon, argues against the formation of the unstructured zircon as a result of diffusional resetting of older grains. Results are given in Table 9 and Fig. 7h. Two grains containing older cores (grains 21 and 43) were analysed (both cores and rims) for REE by LA-ICP-MS (Fig. 8). The cores and the rims of both grains are enriched in heavy REE (HREE) and display negative Eu anomalies. The cores of both grains are more enriched in REE than the rims. Both the REE distribution and geochronology are consistent with the interpretation of two zircon generations. Almost complete dissolution of older grains (defined by the eight older discordant analyses) with variable degrees of Pb loss during a subsequent event is inferred. The morphology and chemistry of the majority of the zircon grains indicate that the c. 540 Ma age dates a metamorphic event. The upper intercept age (c. 1044 Ma) is difficult to interpret because of the imprecision in the estimate. It is, however, within error of ages obtained on a number of other samples (see below), which are interpreted as reflecting an earlier metamorphic event.
|
Tabular pre-tectonic gneissic granite (WBSV073)
The tabular pre-tectonic gneissic granite bodies are intrusive into the paragneiss sequence of the Rootshorga Formation (Groenewald et al., 1995
; Board, 2002). The tabular orthogneiss is intensely migmatized and homogeneous gneissic granite without significant leucosome development was sampled from the Rootshorga massif. The prismatic zircon grains in this rock display rounded terminations, in places leading to an ovoid morphology. Oscillatory growth zonation is developed within all of the zircon grains (Fig. 6b). Dark unzoned zircon overgrowths are commonly developed along the outer margins of the grains and at grain terminations, and are discordant to the growth zonation. Results are given in Table 9 and Fig. 7b.
Chondrite-normalized REE patterns obtained for the magmatic core (207Pb–206Pb age = 1055 ± 27 Ma; 1) and overgrowth (207Pb–206Pb age = 555 ± 20 Ma; 1) of grain 33 (Fig. 8c) display a HREE enrichment (YbN/LaN = 308 and 544, respectively) and strong negative Eu anomalies, but differ in terms of the degree of REE enrichment relative to chondrite. The core is more enriched in REE than the overgrowth. The magmatic cores and secondary overgrowths are consequently interpreted as two generations of zircon on the basis of the textural, chemical and geochronological features presented.
The low Th/U ratios of the secondary overgrowths (Table 9) and their lower REE content suggest that they crystallized during high-grade metamorphism (Williams & Claesson, 1987; Hoskin & Black, 2000). The age of the oldest zircon population (1072 ± 10 Ma) is considered to represent the age of crystallization of the pre-tectonic tabular granite, whereas the 565 ± 11 Ma age is interpreted as dating a metamorphic event. The c. 996 Ma age is possibly of no geological significance and is due to the probe spot on a very thin rim having overlapped with the older core.
Syn-tectonic leucosome domains
Syn-tectonic leucosomes are developed within all pre-tectonic rocks of the southern H.U. Sverdrupfjella. Although two generations were recognized at Skarsnuten, only one generation was found in the Rootshorga and Hamrane groups of nunataks. Four samples of the older generation of leucosome (WBSV071, WBSV074, WBSV113 and WBSV114) and one sample of the younger generation (WBSV116) were analysed by SHRIMP II (Electronic Appendix 2).
Sample WBSV071
Zircon in this semipelitic paragneiss-hosted syn-tectonic leucosome within a ductile shear zone with a top-to-NW shear sense from northern Rootshorga has a complex structure as revealed by CL images, with three types being recognized (Fig. 6c). Older cores displaying magmatic oscillatory compositional zoning are preserved. These are overgrown by weakly cathodoluminescent rims that are discordant to the growth zonation in the cores. A second type of zircon overgrowth is indicated by the presence of unzoned zircon embayments that are characterized by a slightly brighter CL than the first overgrowth type. These embayments are generally developed along the outer margins of the grain and at grain terminations and cut across the first overgrowth type.
The results obtained from 24 analyses of 21 zircon grains are plotted on a U–Pb concordia diagram (Fig. 7c) and summarized in Table 9. The textural, chemical and geochronological characteristics of the zircon grains in this sample indicate three zircon generations. By analogy with sample WBSV073, the low Th/U ratio (<0·1) of the two overgrowth types is interpreted as indicative of a metamorphic origin. The fact that the syn-tectonic leucosome cuts both generations of pre-tectonic felsic intrusives indicates that they must be younger than c. 1072 Ma. Consequently, the older cores are interpreted as resorbed portions of zircon grains inherited from the host paragneiss. The two generations of zircon overgrowth indicate disturbance of the U–Pb isotopic system during two periods of metamorphism, at c. 1035 Ma and c. 499 Ma. The discordant nature of the younger metamorphic overgrowth suggests Pb loss and therefore the c. 499 Ma age is, at best, a minimum age for this generation. As zircon with metamorphic chemical characteristics can be precipitated from metamorphic fluids and partial melts (Williams et al., 1996; Roberts & Finger, 1997), the leucosome domain may either be a Grenvillian body that was reworked during Pan-African metamorphism or a Pan-African domain that contains inherited zircon grains in which relics of an older metamorphic event are recorded.
Sample WBSV074
Zircon grains from this foliation-parallel leucosome within a tabular pre-tectonic orthogneiss from northwestern Rootshorga produce dark, featureless CL images, although compositionally zoned cores characterized by a brighter CL are present in a few cases (Fig. 6d). Unstructured zircon cuts across the growth zonation displayed by the cores and contains isolated remnants of this zonation in some grains.
The results of 28 analyses on 19 zircon grains are presented on a U–Pb concordia diagram (Fig. 7d) and in Table 9. The textural, geochronological and chemical characteristics indicate two generations of zircon growth, and the low Th/U ratio of the overgrowths is consistent with a metamorphic origin. The c. 1153 Ma age is similar to that of the magmatic zircon grains of the oldest granitic intrusive (WBSV065) and the inherited zircon cores of sample WBSV071. The c. 1054 Ma age for the compositionally zoned zircon cores of this sample is within error of the age of the magmatic zircon from the tabular orthogneiss (sample WBSV073) that hosts the leucosome domain from which this sample was collected. Consequently, the older magmatic zircon cores are interpreted as portions of grains inherited from the host gneiss. As the majority of the zircon crystals in this sample consist entirely of unstructured zircon with metamorphic chemical characteristics, except for the few grains in which older cores are preserved, and as no meaningful ages between c. 1054 Ma and c. 515 Ma were determined, the latter is interpreted as dating zircon growth during crystallization of the leucosome.
Sample WBSV113
Leucosome that is parallel to S2 and up to 60 cm wide contains zircon grains that display a homogeneous dark CL image (Fig. 6e). Because of their highly metamict nature, analyses were limited to clear undamaged parts of individual grains.
The results of 13 analyses on 11 grains are presented in Fig. 7e and Table 9. The older population is statistically indistinguishable from the oldest zircon overgrowth of sample WBSV071, and is interpreted as recording a metamorphic event at c. 1032 Ma. The younger population is considered a product from a subsequent metamorphic event at c. 503 Ma that induced partial Pb loss in some of the older zircon grains. The leucosome domain may therefore either be a syntectonic body of Grenvillian age that was reworked during a Pan-African tectono-thermal episode or a wholly Pan-African body that contains relics of an earlier metamorphic event.
Sample WBSV114
This sample comes from the same outcrop as the previous one, but represents a very thin leucosome vein parallel to S2. Most of the zircon grains in this sample are characterized by a very dark CL and zoning is not obvious or is obscured by radiogenic damage (Fig. 6f). In some grains relics of older cores were identified by their bright CL. Some of the cores display a weak oscillatory zonation that is truncated by the darker overgrowths, whereas others exhibit a patchy zonation, suggesting that they have been extensively recrystallized.
The results of 13 analyses on 10 relatively clear zircon grains are plotted on Fig. 7f and summarized in Table 9. A 207Pb–206Pb age of c. 1146 Ma obtained on cores is within error of that determined for the oldest granitic body (WBSV065) and therefore these cores are interpreted as inherited grains. A relatively concordant age of close to 1090 Ma obtained on cores is difficult to interpret. It might date a geological event, but it might also represent a ‘mixed’ age, because of a very low Th/U ratio and patchy zonation under CL. The 207Pb–206Pb age of c. 534 Ma obtained on overgrowths is interpreted as dating zircon growth during crystallization of the leucosome.
Sample WBSV116
This sample comes from the same outcrop as the previous two, but represents a younger leucosome generation that occurs in en echelon tension gashes defining syn-D2 top-to-NW shearing. The zircon grains display a complex internal structure under CL with three types being recognized: older cores are present in some grains and two types of zircon overgrowth can be distinguished (Fig. 6g). The older cores display oscillatory compositional zonation indicating that they crystallized out of a melt. The one type of zircon overgrowth is characterized by bright CL and occurs as rims to the older zircon cores, as discrete structureless grains and as cores to those grains that are rimmed by the second overgrowth type. Overgrowths of this first type truncate the compositional banding of the older cores. The second type of zircon overgrowth is characterized by dark CL and occurs as rims and as embayments that are locally discordant to the bright CL overgrowths. In some grains zircon with bright CL forms a rim to structureless zircon cores characterized by a dark CL, suggesting that the two types of zircon overgrowth may be coeval.
Results obtained on 22 analyses from 18 grains are plotted on a U–Pb concordia diagram (Fig. 7g) and summarized in Table 9. Whereas the textural evidence indicates three types of zircon, the chemical and isotope data are consistent with the zircon overgrowths having formed penecontemporaneously during the crystallization of the leucosome. Consequently, the leucosome is considered to be of Pan-African age.
U–Pb SHRIMP dating of monazite
Monazite occurs in pelitic paragneiss as clear, yellow, angular to rounded, anhedral grains, ranging in size from 130 to 390 µm. Although BSE imaging revealed that the majority of the monazite grains are homogeneous, small dark irregularly shaped cores are present in a few grains as well as diffuse areas with a brighter signal (Fig. 9a). SHRIMP I was used to obtain 20 analyses from 14 grains (Fig. 10a). Three analyses of small dark cores (8.1, 15.1, 20.1) yielded 207Pb–206Pb ages (1 errors) of 1014 ± 8 Ma (10% discordant), 1077 ± 8 Ma (4% discordant) and