Dumond 2015

J. metamorphic Geol., 2015, 33, 735–762 doi:10.1111/jmg.
12150
Monazite as a monitor of melting, garnet growth and feldspar

recrystallization in continental lower crust
G. DUMOND,1* P. GONCALVES,2 M. L. WILLIAMS3 AND M. J. JERCINOVIC3
1
Department of Geosciences, University of Arkansas, Fayetteville, AR 72701, USA ([email protected])
2
UMR 6249 Chrono-Environnement, Universit e de Franche-Comt e , 16 route de Gray, 25030 Besancß on, France
3
Department of Geosciences, University of Massachusetts, Amherst, MA 01003, USA
ABSTRACT Monazite is a common accessory phase in felsic granulite ribbon mylonites exposed in the Upper
Deck domain of the Athabasca granulite terrane, western Canadian Shield. Field relationships, bulk
rock geochemistry and phase equilibria modelling in the Na2O–CaO–K2O–FeO–MgO–Al2O3–SiO2–
H2O–TiO2–Fe2O3 system are consistent with the garnet-rich rocks representing the residual products
of ultrahigh temperature melting of biotite-bearing paragneisses driven by intraplating of mafic
magma in continental lower crust. The c. 2.64–2.61 Ga Y-rich resorbed monazite cores included in
garnet are interpreted as relicts of detrital grains deposited on the Earth’s surface after c. 2.61 Ga.
Yttrium-poor monazite domains in garnet are depleted in Sm and Gd and linked to fluid-absent melt-
ing of biotite + plagioclase + quartz sillimanite during a prograde loading path from ≤0.8 to
≥1.4 GPa. The c. 2.61–2.55 Ga Y-depleted, Th-rich monazite domains crystallized in the presence of
garnet + ternary feldspar orthopyroxene + peraluminous melt. The c. 2.58–2.52 Ga monazite rims
depleted in Th + Ca and enriched in Eu are linked to localized melt extraction synchronous with
growth of high-pressure (HP) grossular-rich garnet at the expense of plagioclase during crustal thick-
ening, culminating at >950 °C. Re-heating and dextral transpressive lower crustal reactivation at
c. 1.9 Ga resulted in syn-kinematic growth of (La + Ce)-enriched monazite and a second generation
of garnet, concurrent with recrystallization of feldspar and orthopyroxene at 1.0–1.2 GPa and 600–
700 °C. Monazite grains in this study are marked by positive Eu-anomalies relative to chondrite. A
direct link is implied between Y, Sm, Eu and Gd in monazite and two major phases in continental
lower crust: garnet and plagioclase. Positive Eu-anomalies in lower crustal monazite associated with
modally abundant garnet appear to be directly related to Eu-enrichment and depletions of Y, Sm and
Gd that are consequences of garnet growth and plagioclase breakdown during HP melting of peralu-
minous bulk compositions.
Key words: continental crust; garnet; granulite; monazite; partial melting.
The ability to accurately date the processes involved

INTRODUCTION
with weakening and differentiating continental crust
Partial melting is a fundamental mechanism for requires knowledge of the reactions leading to growth
weakening and differentiating continental crust. or dissolution–precipitation of accessory phases dur-
Fluid-absent partial melting of biotite-bearing quart- ing and after partial melting. This study explores
zofeldspathic gneisses, in particular, represents one links between monazite and fluid-absent melting of
important process for generating granitic melts and biotite in continental lower crust, including the
their complementary dense garnetiferous residues production of garnetiferous residues and feldspar
(Clemens & Vielzeuf, 1987; Rudnick, 1992; Vielzeuf recrystallization.
& Montel, 1994; Clemens, 2006; Johnson et al., Three approaches have been applied to under-
2008). Partial melting also leads to strain localization, standing the monazite + melt system. Monazite
melt-enhanced deformation and dramatic changes in solubility during melting has been investigated experi-
crustal rheology (Hollister & Crawford, 1986; Brown mentally (Montel, 1986, 1993; Rapp & Watson, 1986;
& Solar, 1998; Searle, 2013; Yakymchuk & Brown, Rapp et al., 1987; Skora & Blundy, 2012; Stepanov
2014a). Models for crustal-scale channel flow rely on et al., 2012; Xing et al., 2013). Recent theoretical
‘melt weakening’ to facilitate the necessary drop in work has focused on thermodynamic forward
viscosity and bulk rheology for crustal flow (Beau- modelling of common bulk compositions to investi-
mont et al., 2001, 2004; Rosenberg & Handy, 2005). gate the growth and dissolution of monazite in melt-
bearing systems (Kelsey et al., 2008; Yakymchuk &
*This manuscript is dedicated to the memory of Professor J. A. Brown, 2014b). The third approach involves the
Grambling (1953–1993). study of monazite composition and texture in natural
© 2015 John Wiley & Sons Ltd 735

736 G. DUMOND ET AL.
systems, e.g. migmatites, leucogranites and granulites constraints on the evolution of continental lower crust
(Watt, 1995; Pyle & Spear, 2003; Goncalves et al., during partial melting, crustal flow, crustal thickening
2004; Kelly et al., 2006; Rubatto et al., 2006; and strain partitioning.
Dumond et al., 2008, 2010; Crowley et al., 2009;
Martins et al., 2009; Reno et al., 2012; Lederer et al.,
GEOLOGICAL BACKGROUND
2013; Yakymchuk et al., 2015, among others). We
use monazite in high-pressure (HP) peraluminous
Athabasca granulite terrane
felsic granulite paragneisses to investigate deep crus-
tal partial melting and post-melting crustal evolution. The Athabasca granulite terrane is >20,000 km2 of
Monazite is a common accessory phase in felsic exhumed continental lower crust (0.8 to
granulite ribbon mylonites exposed in the >400 km2 >1.5 GPa 30 to >55 km paleodepths) exposed
Upper Deck domain of the Athabasca granulite north of Lake Athabasca, Canada (Fig. 1a; Dumond
terrane, one of the Earth’s largest exposures of conti- et al., 2008; Mahan et al., 2008). The terrane is com-
nental lower crust (>20,000 km2), in the western posed of Archean to Paleoproterozoic mafic and fel-
Canadian Shield. Felsic granulite paragneisses in the sic granulites, orthogneisses, paragneisses and minor
~10 km thick Upper Deck are inferred to represent eclogite bounded on its eastern margin by the
the residual products of ultrahigh temperature 1.85 Ga Legs Lake shear zone (Fig. 1b; Mahan
(UHT) melting of continental lower crust (Snoeyen- et al., 2003, 2006a,b; Mahan & Williams, 2005). Two
bos et al., 1995; Baldwin et al., 2006). High-resolu- main groups of structures have been identified
tion X-ray mapping, in situ high-precision electron throughout the terrane. The first group (D1) includes
probe microanalysis (EPMA), major and trace element sub-horizontal gneissic fabrics, ESE-trending stretch-
geochemistry, and phase equilibria modelling reveal ing and intersection lineations and recumbent isocli-
first-order links between composition, microstructure, nal folds (Dumond et al., 2010; Regan et al., 2014).
Th–U–total Pb monazite geochronology, and felsic These earliest structures are interpreted to have
granulite petrogenesis. These links provide tighter resulted from granulite facies lower crustal flow at
(a) 60 ° (b) Whold

N a
Lake ia
Lake
ke
Athabasca
La
0.9-1.0 GPa
rd
800-900 ºC
1
bi
ow
N Sn
0.76 GPa
North 840 ºC
sz
2
America
LL
Selwyn 1.12 GPa
Lake 795 ºC 1.16 GPa
500 km 102 °W
2 820 ºC
3
sz ?
RAE CL 1.35 GPa Kasba
N
4 800 ºC
(b) 0.9-1.0 GPa 9 0.9-1.0 GPa 6 Lake
0.8 GPa 750-800 ºC
O 800-900 ºC 800-900 ºC
GRsz 0 50 km
Bs 10 1.0-1.1 GPa
0.9 GPa
750 ºC
5
z 0.9 GPa 750 ºC
Explanation P–T point
800 ºC 6 1.3 GPa
11 1.7 Ga Athabasca basin with reference
GRsz
850 ºC
4 c. 1.93-1.85 Ga
Rae domain undifferentiated
N sz
0.5 GPa
600-650 ºC Hearne domain undifferentiated 1.16 GPa
C L 820 ºC 3
7 Paragneiss
L
Athaake U C HEAR Granite, granodiorite orthogneiss c. 2.62-2.50 Ga
sz
bas NE
ca
LL
1.0-1.2 GPa 8 Felsic granulite 1.3 GPa

59 °N
775-860 ºC >1.5 0.9-1.0 GPa 850 ºC

1.2 GPa GPa 750-800 ºC
1.1 GPa Mafic granulite, amphibolite 4
750-800 ºC 12 6 750-850 ºC
13 > 900 ºC 14 105 °W Tonalite gneiss Figure 2
Fig. 1. (a) Location of the Athabasca granulite terrane in North America near the east end of Lake Athabasca. (b) Geological
map of the Athabasca granulite terrane with existing P–T–t data referenced by number (modified after Mahan et al., 2008;
Dumond et al., 2013): (1) Krikorian (2002); (2) Martel et al. (2008); (3) Mahan & Williams (2005); (4) Mahan et al. (2008); (5)
Regan et al. (2014); (6) Williams et al. (2000); (7) Mahan et al. (2003); (8) Williams et al. (1995); (9) Williams & Hanmer (2006);
(10) Kopf (1999); Williams & Jercinovic (2002); (11) Dumond et al. (2010); (12) Kopf (1999); (13) Baldwin et al. (2003); (14)
Snoeyenbos et al. (1995); Baldwin et al. (2004). Circled letters correspond to lithotectonic domains within the East Athabasca
mylonite triangle mentioned in the text: (C) the Chipman domain, (N) the Northwestern domain and (U) the Upper Deck domain.
The rectangle outlines the location of Fig. 2. OBsz, Oldman-Bulyea shear zone; GRsz, Grease River shear zone; CLsz, Cora Lake
shear zone; LLsz, Legs Lake shear zone.
© 2015 John Wiley & Sons Ltd

MONAZITE IN CONTINENTAL LOWER CR UST 737
2.60–2.55 Ga in the Rae cratonic province (Dumond 2011). Rocks in the Upper Deck consist of garnet-
et al., 2010; Regan et al., 2014). Younger generation rich felsic granulite, garnetite and paragneisses that
structures (D2–D5) include SW-dipping gneissic folia- are interlayered with sills of mafic granulite and
tions that are overprinted by upright, open to tight minor eclogite (Fig. 2; Hanmer, 1994, 1997;
folds with localized axial planar and shear foliations, Snoeyenbos et al., 1995; Baldwin et al., 2003, 2004,
and transposition of folds and gneissic fabrics into 2006, 2007, 2015). The felsic granulite gneisses are
sub-vertical to steeply dipping NE-striking mylonitic characterized by a penetrative mylonitic foliation
shear zones (Mahan et al., 2003, 2008; Dumond defined by ribbons of quartz and recrystallized tern-
et al., 2008, 2010, 2013; Martel et al., 2008; Bethune ary feldspar, in association with Grt + Ky + Rt
et al., 2013; Regan et al., 2014). All D2–D5 structures Sil Opx and minor plagioclase (mineral abbrevia-
are attributed to sub-horizontal shortening related to tions after Bucher & Frey, 2002). The typical white
the Taltson and Wopmay orogenies (to the west) and colour and mylonitic character of these rocks, infor-
the Trans-Hudson orogeny (to the east) during the mally called the ‘white gneisses’, were inferred by
interval of 1.94–1.80 Ga (Dumond et al., 2008, 2013; Snoeyenbos et al. (1995) to be analogous to the
Bethune et al., 2013; Regan et al., 2014). ‘Weiss-stein’ granulites from the type locality in the
The East Athabasca mylonite triangle (EAmt; Bohemian Massif (Weiss, 1803; O’Brien & R€ otzler,
Tantato domain of Slimmon, 1989) is the most thor- 2003). Minimum peak metamorphic conditions for
oughly studied portion of the Athabasca granulite the felsic granulite gneisses are in the order of
terrane (Hanmer, 1994, 1997). The triangle is 1000 °C at ≥1.5 GPa, assuming equilibrium between
divided into three lithotectonic subdomains: Chip- ternary feldspar and kyanite (Snoeyenbos et al.,
man, Northwestern and Upper Deck (C, N and U 1995).
in Fig. 1b; Hanmer et al., 1994; Williams & Han- Geochronology by ID-TIMS (zircon and monazite)
mer, 2006; Williams et al., 2009; Mahan et al., and EPMA (monazite) on the felsic granulite gneisses
04G-097
Upper Deck, Athabasca granulite terrane 22 S1 foliation/enveloping surface 01SZ90 sample
S2 foliation upright F2 antiform
FOOTWALL
eclogite, c. 1.9 Ga Robillard 72

HANGING
garnet clinopyroxenite granite (Hanmer 12 streching/intersection lineation upright F2 synform

WALL
mafic granulite et al., 1994) dextral transpressive shear zone overturned F2 synform
N felsic granulite, garnetite c. 2.61-2.53 Ga
Godfrey granite normal ductile detachment zone overturned F1 antiform
felsic granulite protolith (Baldwin et al., 2003) (block on downthrown side)
0 10 km
= S1 (n = 27)
Honsvall (a) Godfrey Bay detachment zone
N Lake
(b) Algold Bay N = F2 β-axis Currie
Lake = S1 (n = 30)
shear zone 10 N
9 18 = L1 (n = 27)
= S2 (n = 53)
= L2 (n = 61) 23 eclogite 18
25 24
65 22
70 32 30 12 25 d
d 85 07G-030-H2A u UTM
24
Melby u 07G-014A Axis 01SZ90 Northing
Tederoff Lake Lake 6580
Lake
Algold UTM 460
Bay Kennard Hardie Easting
Lake F2 Lake
F2
6 05G019B 22
05G-047C A 15 21 A’
65 2 05G-026B 72 12 28 21
19 F2 48 54 22
04G-09712 64
53 38 12 1 33 eclogite 81 F1 28
19 5 N
04G-090A 45 78 16 18
Sucker 16
on e
Bay 37 13
r z ak
e
50 18 7105G-051B 17 d
ea L
23
N Stony Rapids,
= S1 (n = 17) 05G-036 F
sh ora
21 Lake Athabasca
21 1 = S1 (n = 4) Saskatchewan, Canada
= S2 (n = 95) u
C
(c) = S2 (n = 28)
= F3 β-axis = L2 (n = 25)
A A’
= 209 m above
sea level
(mean elevation of
5 km coastal exposure)
Horizontal = Vertical
Scale Scale
(d)
= Folded L2 (n = 96)
Fig. 2. Simplified geological map and cross-section A–A0 of the Upper Deck domain with sample locations (this study; Hanmer,
1994; Baldwin et al., 2003, 2004, 2006). Lower hemisphere equal area projection stereonets are illustrated for structural data from:
(a) the Godfrey Bay detachment zone along the northeast boundary of the domain; (b) the Algold Bay shear zone along the
northwest boundary of the domain; (c) Kennard Lake; and (d) Sucker Bay and the western portion of the Upper Deck. Open
circles are poles to S1 and filled circles are poles to S2. Diamonds correspond to stretching and intersection lineations.

was used to infer an episode of melting and garnet veined internally and along their contacts by leuco-
growth at c. 2.55 Ga (Baldwin et al., 2006). Mafic somes (Fig. 3d). These observations are consistent
granulite gneisses hosted by the felsic granulite with partial melting of the paragneisses driven by
gneisses record temperatures of 890–960 °C and emplacement of the igneous protolith to the mafic
pressures ranging from 1.3 to 1.9 GPa with two granulite sill complex.
populations of U–Pb ID-TIMS zircon dates at The migmatitic paragneisses in Sucker Bay occur
c. 2.55–2.52 and c. 1.9 Ga (Baldwin et al., 2003). in gradational contact with orthopyroxene-bearing
Eclogite HP metamorphism (1.6 GPa and 750 °C) in felsic granulite gneisses. These gneisses contain tern-
the Upper Deck domain has been interpreted at ary feldspar + Grt + Opx + Qtz + Rt + Ilm + Po.
c. 1.9 Ga, based on U–Pb dates obtained in situ via Elongate feldspar and quartz ribbons define a pene-
SHRIMP for zircon grains associated with recrystal- trative ribbon mylonite gneissic fabric (Fig. 3e). Gar-
lized clinopyroxene in the eclogite gneisses (Baldwin net + orthopyroxene-bearing leucogranites derived
et al., 2004). However, high-precision single-grain from the orthopyroxene-felsic granulites locally cross-
ID-TIMS data for multiple zircon grains from the cut the mafic granulites (Fig. 3f). These felsic gneisses
same rock produced a well-defined discordia between progressively grade along strike and up section into
c. 2.54 and 1.90 Ga (sample 01SZ40B in Baldwin orthopyroxene-absent felsic granulite gneisses com-
et al., 2004). posed of ternary feldspar + Grt + Ky Sil + Qtz +
Rt + Gr Pl (Fig. 4), similar in character to those
previously described by Snoeyenbos et al. (1995) and
FIELD RELATIONSHIPS IN THE UPPER DECK
Baldwin et al. (2006).
DOMAIN
Three main lithologies characterize the Upper Deck
Upper Deck domain boundaries and internal structure
domain: (i) sills of abundant mafic granulite and
minor eclogite gneisses, (ii) felsic granulite gneisses The >400 km2, ~10 km thick Upper Deck domain
and (iii) previously unrecognized biotite-bearing structurally overlies the northwestern domain along
paragneisses interpreted to be the protolith to the a gently SW-dipping, top-to-the-SW ductile shear
felsic granulite gneiss (Fig. 2), discussed below. New zone (Hanmer et al., 1994, 1995), here named the
mapping, combined with data from Hanmer (1994), Godfrey Bay detachment zone (Fig. 2a). Kinematics
provides structural context for rocks of the Upper defined by recrystallized asymmetric porphyroclasts
Deck and its tectonic boundaries with the other sub- of orthopyroxene and feldspar define a normal,
domains of the EAmt (Figs 1b & 2). Rocks in the down-dip to dextral oblique-slip sense of shear
immediate footwall of the Upper Deck include high- (Hanmer et al., 1995). Penetrative ribbon mylonite
strain orthogneisses represented by the c. 1.9 Ga fabrics of the Godfrey Bay detachment zone merge
Robillard granite pluton (Hanmer et al., 1994) and westwards into the NE-striking, moderately SE-dip-
the c. 2.61–2.53 Ga Godfrey granite pluton (Fig. 2; ping Algold Bay shear zone (Fig. 2). Dextral,
Baldwin et al., 2003). SE-side up transpressive kinematics along gently
NE-plunging stretching lineations in the shear zone
are defined by mm- to cm-scale Opx r-clasts in
Lithologies in the western Upper Deck domain
Algold Bay and map-scale C–S foliation trajectories
Mafic granulite sills occur throughout the ~10 km in the vicinity of Tederoff Lake (Fig. 2b). The tec-
thick Upper Deck domain as <1 cm to >1 km-thick tonic significance of these two shear zones is the
lenses composed of Grt + Cpx + Qtz Pl Hbl subject of a separate study. The boundary of the
gneiss (Fig. 2; see also Baldwin et al., 2003). Folded Upper Deck domain is cut by the NE-striking, stee-
and transposed layers of minor eclogite (locally ply NW-dipping Cora Lake shear zone (Fig. 2;
>10 m-thick and >10 km in strike-length) are exposed Mahan et al., 2008; Regan et al., 2014). The 3–5 km
north of Axis Lake and as highly retrogressed bou- wide shear zone is defined by ultramylonite fabrics
dined lenses (<1 to 2 m-thick) northeast of Sucker in granulites that record a sinistral normal oblique-
Bay (Fig. 2; see also Snoeyenbos et al., 1995; Bald- slip history along a moderately SW-plunging
win et al., 2004, 2007, 2015). The mafic granulite sills stretching lineation at c. 1.88 Ga during the early
are hosted by lithologically heterogeneous stromatic stages of exhumation of the Athabasca granulite
migmatite interpreted to be derived from paragneisses terrane (Regan et al., 2014).
in the southwestern portion of the Upper Deck near All major lithologies in the Upper Deck domain
Sucker Bay (Figs 2 & 3a–c). The paragneisses were affected by a D1 deformation event defined by
become more migmatitic with increasing proximity to sub-horizontal to variably gently dipping penetrative
the mafic granulite sills, where the paragneisses con- ribbon mylonite fabrics (S1) (Figs 2b–d, 3a & 4a,b).
tain patchy dm-scale neosomes (Fig. 3c) that transi- Fabrics associated with D1 are locally defined by
tion into penetratively deformed layers of garnetite blades of kyanite and layers of garnetite + leuco-
and leucosomes containing garnet and orthopyroxene some (Fig. 4a,b). These observations are most
(Fig. 3c–e). Sills of mafic granulite are intensely consistent with migmatite generation occurring

(a) (b)
(c) (d)
(e) (f)
Fig. 3. Outcrop photos of field relationships between the felsic granulite paragneiss protolith, mafic granulite gneisses and Opx-
bearing felsic granulite gneisses in the western Upper Deck domain near Sucker Bay in Fig. 2. Sample locations are noted by arrows
with sample numbers. (a) Outcrop of metre-scale, mafic granulite sills hosted by Bt-bearing paragneiss protolith. All gneisses in the
outcrop are penetratively deformed by a gently dipping S1 ribbon mylonite fabric. (b) Close-up view towards east of mafic granulite
sills in the background of picture in (a). Note the contact zone between the Bt-bearing paragneiss protolith and mafic granulite sills
(dipping <25° to viewer’s right) and the abundance of leucosome and garnetite pods in the contact zone. (c) Biotite-paragneiss
protolith near mafic granulite with Grt + Opx neosome. (d) Close-up view of Opx-bearing leucosome layers and cross-cutting veins in
the mafic granulite sills in (b). (e) Characteristic ribbon mylonite fabric in the Opx-bearing felsic granulite defined by ribbons of Qtz +
ternary feldspar and elongate clasts of Opx + Grt. (f) Leucogranite dyke derived from Opx-felsic granulite that cross-cuts a mafic
granulite sill and has apophyses that intrude parallel to layering in the sill.

(a) (b)
(c) (d)
Fig. 4. Outcrop photos of aluminosilicate-bearing (Opx-absent) felsic granulites in the western Upper Deck domain near Kennard
Lake (see Fig. 2 for location). (a) and (b) Outcrop and sketch of Grt + Ky + Sil felsic granulite ribbon mylonite. Note the early
NW-striking S1 fabric defined by leucosomes + layers of garnetite with an enveloping surface similar in orientation to S1 fabrics
observed in Sucker Bay (Fig. 3a). The outcrop is overprinted by a penetrative, steeply NW-dipping S2 cleavage. Note location of
sample 05G-019B for this study. (c) Outcrop of fine-grained felsic granulite with a penetrative D2 ultramylonite fabric defined by
ribbons of quartz and feldspar. Note apparent grain-size reduction in garnet. (d) View parallel to the stretching lineation and
perpendicular to the foliation (XZ-plane of strain). Note mantled garnet d-clast with recrystallized tails of plagioclase defining
dextral sense of shear.
synchronous with the S1 sub-horizontal fabric.

MICROSTRUCTURE, PETROGRAPHY AND
Deformation during D2 involved transposition of
MINERAL COMPOSITIONS
these gneissic mylonite fabrics into overturned tight
to isoclinal folds (F2) associated with NE-striking Nine samples were chosen from a suite of over 70
steeply dipping to sub-vertical foliations (S2) and samples of felsic granulite and paragneiss protolith
well-developed variably plunging intersection and rocks collected along a 20 km-long transect in the
stretching lineations (L2; Figs 2b–d & 4b). The D2 western portion of the Upper Deck (between
event is also associated with zones of sub-vertical Sucker Bay and Kennard Lake) and in the vicinity
NE-striking S2 ultramylonite fabrics with gently of Honsvall Lake (as identified in Fig. 2; for refer-
NE-plunging stretching lineations (L2) defined by ence, felsic granulite sample 01SZ90 from Baldwin
ribbons of quartz and feldspar (Fig. 4c). Dextral et al., 2006; is plotted as a star in Fig. 2). Whole-
kinematics observed parallel to L2 and perpendicu- rock geochemistry for eight of the samples is
lar to S2 are defined by mantled garnet and summarized in Table S1 and discussed below. Three
orthopyroxene porphyroclasts (Figs 3e, 4d & 5c). samples were chosen for detailed microanalysis.
All S1 and S2 fabrics, in addition to L2 lineations Sample 04G-097 is an orthopyroxene-bearing felsic
and F2 fold axes, were subsequently folded during granulite ribbon mylonite containing a penetrative
a D3 event (Fig. 2d). gently SE-dipping S1 foliation and SW-plunging

(a) (b)
(c) (d)
(e) (f)
Fig. 5. Microstructures and X-ray compositional maps for sample 04G-097. (a) Photomicrograph (plane-polarized light) and Mg
Ka map of peak assemblage in Opx-bearing felsic granulite: Opx + Grt1 + Qtz + Rt + ternary feldspar. Ilmenite is a matrix phase
that was not observed in garnet. (b) Large anti-perthite (ternary feldspar) porphyroclast with resorbed Ti-rich biotite inclusion. (c)
High-resolution X-ray maps of recrystallized orthopyroxene for Al Ka and Mg Ka. Note corona of Grt2 around orthopyroxene in
Mg Ka map and diffuse high-Al core in Al Ka map. (d) Large poikilitic porphyroblast of Grt1 with inclusions of Bt + Pl + Qtz +
Mnz + Zrn + Rt. (e) High-resolution Ca Ka X-ray map of garnet depicted in (d). Note recrystallized ternary feldspar matrix above
garnet. Garnet porphyroblasts are characterized by low-grossular cores surrounded by high-grossular annuli. Note location of
monazite 04G-097-2-m1 in high-grossular Grt1 (illustrated in Fig. 12a). (f) Backscattered electron image of poly-phase inclusion of
Bt + Pl + Qtz + Kfs + Zrn + Rt in high-grossular Grt1 annulus.
stretching lineation defined by ribbons of quartz + of the felsic granulites, and texturally link these
feldspar. This sample was chosen to constrain peak results to monazite composition and geochronology.
metamorphic conditions, model the P–T evolution As mentioned previously, the orthopyroxene-felsic

granulites gradationally transition into Sil + Ky-bear- lar-rich garnet domains and include K-feldspar +
ing and garnet-rich felsic granulites and garnetite that quartz + plagioclase + biotite + rutile + zircon
lack orthopyroxene. An oriented sample (05G-019B) (Fig. 5f). Plagioclase (An27) also occurs as resorbed
of the kyanite-rich felsic granulite was collected from grains with biotite as inclusions in grossular-rich gar-
the outcrop depicted in Fig. 4a,b to constrain the net (Fig. 5d,e). The second occurrence of garnet con-
P–T conditions and age of garnet growth and partial sists of small (0.10–0.30 mm) grossular-rich grains
melting indicated by the leucosomes and garnetite (up to Grs16) that, with quartz, define coronae
layers that define the early S1 fabric. Metre-scale, around dynamically recrystallized orthopyroxene por-
sub-vertical ultramylonite zones are abundant in the phyroclasts (Fig. 5a,c). Garnet grains that define the
kyanite-bearing felsic granulites (Fig. 4c). The third corona are typically poikiloblastic with inclusions of
sample (05G-026B) is from a D2 dextral ultramylonite plagioclase and quartz.
zone to constrain the timing of D2 strain and the Microstructural observations support growth of
potential relationship between monazite growth and Grt1, orthopyroxene and ternary feldspar by a fluid-
feldspar recrystallization. absent biotite melting reaction of the type Bt + Pl +
High-resolution X-ray mapping of silicate phases Qtz ? Opx + Grt + Kfs + melt (e.g. Vielzeuf & Mon-
was carried out prior to analysis with the CamecaÒ tel, 1994). By contrast, the Grt2 corona texture
SX50 EPMA at the University of Massachusetts- around orthopyroxene grains (Fig. 5a,c) may have
Amherst. Maps were generated at 15 kV and 150– occurred according to a reaction of the type Opx +
200 nA with 4–10 lm pixel step sizes and 50–100 ms Pl ? Grt + Qtz.
dwell times. Quantitative analyses, guided by the
X-ray maps, were collected at 15 kV and 20 nA with
Sillimanite + kyanite-bearing felsic granulite (sample 05G-
a focused beam for garnet and orthopyroxene and a
019B)
defocused beam (5 lm diameter) for biotite and pla-
gioclase (Table S2). Count times were 20 s on peak Major phases in the aluminosilicate-bearing felsic
and 10 s on background. Calibrations were made granulite gneisses include garnet, ternary feldspar,
using common natural and synthetic standards. quartz, kyanite, sillimanite, with minor plagioclase
and biotite (Fig. 6). Accessory phases include abun-
dant rutile + monazite + zircon + graphite + apatite.
Orthopyroxene-bearing felsic granulite (sample 04G-097)
Fluorapatite occurs only as rounded grains (up to
Major phases in the orthopyroxene-bearing felsic 150 lm) included in garnet and is completely absent
granulites are garnet, orthopyroxene, ternary feldspar from the matrix (Fig. 6a).
(as perthite and anti-perthite), biotite, quartz and Elongate quartz ribbons, garnet aggregates, blades
minor plagioclase (Fig. 5a–c). The main penetrative of kyanite + sillimanite (up to 1 mm long) and
S1 fabric is defined by ribbons of quartz, aggregates recrystallized ribbons of ternary feldspar define the
of garnet, and dynamically recrystallized orthopyrox- sub-horizontal to gently dipping S1 fabric. The
ene, ternary feldspar and biotite. Zircon, monazite, upright S2 foliation is locally defined by fractured
rutile, ilmenite and pyrrhotite are abundant as matrix aggregates of resorbed garnet with fractures filled by
accessory phases, and all occur as inclusions in garnet kyanite + sillimanite + quartz + rutile. The textural
(Figs 5d–f & S1), with the notable exception of ilme- relationship between kyanite and sillimanite, as
nite (Fig. 5a). Ternary feldspar (anti-perthite) occurs observed in the matrix, remains equivocal. However,
as large porphyroblasts that locally contain resorbed tens of micron prismatic grains of sillimanite occur as
Ti-rich biotite inclusions (up to 6.25 wt% TiO2; inclusions in garnet that may have nucleated prior to
Table S2) similar to those preserved in garnet resorbed kyanite inclusions in garnet rims (Fig. 6a).
(Fig. 5b). Biotite also occurs as a minor matrix phase Recrystallized ribbons of quartz + K-feldspar +
(Fig. 5a). plagioclase + kyanite + sillimanite also define S2.
Orthopyroxene porphyroclasts (3–10 mm) occur as Dextral oblique-slip kinematics for D2 strain are
fractured grains with undulose extinction and have defined by isolated garnet porphyroclasts mantled by
diffuse high-Al cores (up to 4.48 wt% Al2O3) and recrystallized quartz and plagioclase (Figs 2c, 4d &
low-Al rims (1.08 wt% Al2O3; Fig. 5c; Table S2). S3).
Garnet occurs in two microstructural domains. The Garnet (up to 50% by mode) varies from inclu-
first occurrence consists of large poikiloblastic grains sion-poor to poikilitic with inclusions of sillimanite,
(1–3 mm) in leucocratic layers with ternary feldspar, kyanite, rounded quartz, monazite, zircon, rutile,
quartz and biotite (Fig. 5d,e). The garnet graphite and rare rusty brown anhedral biotite
poikiloblasts have grossular-poor cores (Grs4) and (Figs 6a & S2). Garnet cores contain crystallographi-
grossular-rich annuli (up to Grs13) near the rims cally oriented needles of rutile, inferred to represent
(Fig. 5e; Table S2). The grossular-rich domains are exsolution from a more Ti-rich garnet (Snoeyenbos
characterized by the occurrence of Ti-rich biotite et al., 1995). Polyphase inclusions near garnet rims
inclusions (up to 6.07 wt% TiO2; Table S2). Poly- are common, and consist of kyanite + K-feldspar +
phase inclusions in garnet are common in the grossu- plagioclase + quartz + rutile + zircon (see also

(a) 400 µm (b) 200 µm (c)

Ca Kα BSE
Pl Grt1 Grt2
Sil 700 7
Mg (Pyr)
maps below
high
maps below
600 6
500 5
Ca (Grs)
P (ppm)
Ky 400 40 4
Pl
21
300 30 3
low
1
Pl
Plagioclase (An)
24 1 21 200 20 2
22
100 10 1
20
0 40 80 0
Distance (μm) Mnz 0 50 100 150 200 250 300 350
Pl XZ-2-m24
Sil Pl Distance (µm)
Mnz
Fig. 13d XZ-m6
Fig. 13f
Ap
Pl
200 µm 200 µm
Ca Kα Ca Kα
P Kα P Kα
Grt2
Grt1 Fig. 6. (a) and (b) High-resolution X-ray maps and
backscattered electron (BSE) image of garnet in Ky +
(c) Traverse
Pl Sil-bearing felsic granulite 05G-019B illustrating setting of

monazite inclusions depicted in Fig. 13. Dashed white line in
phosphorous maps marks the boundary between high-
Grt2 phosphorous Grt1 cores and low-phosphorous Grt2 rims
discussed in the text. (c) Quantitative EPMA traverses across
Grt1 the boundary between Grt1 cores and Grt2 rims. Grt1 includes
200 µm Pl 200 µm low-Grs garnet cores (Grs3–4) that grade into high-Grs annuli
(Grs7–11).
Snoeyenbos et al., 1995; Baldwin et al., 2006). Table S2), recrystallized ternary feldspar + quartz,
Rounded polyphase inclusions of biotite + quartz (up and/or intergrowths of kyanite rutile.
to 0.24 mm) are less commonly present in garnet. Observations are most consistent with growth of
Biotite inclusions are rich in Ti (up to 5.42 wt% Grt1, defined by low-grossular cores and high-grossu-
TiO2; Table S2). Some quartz grains included in gar- lar annuli, via a fluid-absent partial melting reaction
net contain inclusions of either graphite or rutile of the type Bt + Pl + Sil + Qtz ? Grt + Kfs + melt
(Fig. S2). Two occurrences of garnet were identified (with Ky present at higher pressures; Tropper et al.,
based on trace element maps of phosphorous 2005). The association and local intergrowth of low-
(Fig. 6a,b), an element interpreted to be relatively grossular Grt2 with secondary plagioclase (Fig. 6b)
unaffected by retrograde diffusion (Spear & Kohn, may represent a low-P assemblage that grew some-
1996). The two occurrences of garnet are defined by time after decompression and partial breakdown of
high-phosphorus (200–300 ppm) cores (Grt1) and the high-grossular Grt1 annuli (Snoeyenbos et al.,
low-phosphorous (<150 ppm) rims (Grt2; Fig. 6a,b). 1995).
The sharp contact separating cores from rims is coin-
cident with the occurrence of an ~50 lm wide high-
Felsic granulite ultramylonite (sample 05G-026B)
grossular annulus (Grs7–8, locally up to Grs11 and
Alm57Pyp32), comparable with the grossular-rich Much of the felsic granulite in the western Upper
annuli observed in the orthopyroxene-bearing felsic Deck domain is over-printed by metre-scale, steeply
granulite 04G-097 (cf. Fig. 5e with Fig. 6a,b). The dipping D2 ultramylonite zones (Fig. 4c). Sample
interior margin of the annuli (towards the garnet 05G-026B is a monazite-bearing felsic granulite from
core) locally coincides with small (20–30 lm) pris- an ultramylonite zone that contains a penetrative,
matic inclusions of sillimanite (Fig. 6a). The exterior steeply NW-dipping foliation and gently NE-plunging
margin of the annuli (towards the garnet rim) locally stretching lineation defined by ribbons of ternary
coincides with resorbed inclusions of kyanite feldspar, plagioclase and quartz (Figs 4c & S3). Frac-
(Fig. 6a). The inner core of Grt1 and the outer rim tured porphyroclasts of garnet are mantled by recrys-
of Grt2 display similar low-grossular compositions tallized ribbons of ternary feldspar, plagioclase and
(Grs3–4, Alm59–61Pyp35–36; Fig. 6c). The low-phospho- quartz that define r-type geometries and dextral,
rous Grt2 rims are locally mantled and intergrown oblique-slip kinematics (Fig. S3). Elongate monazite
with plagioclase (An20 cores and An23 rims; Fig. 6a,b; grains are aligned sub-parallel to S2, in association

with ribbons of ternary feldspar and quartz (HREE) enrichments and slightly positive Eu-anoma-
(Fig. 14a). lies consistent with garnet and feldspar accumulation
respectively (Fig. S5). The Opx-bearing leucogranite
dyke, in particular, exhibits a well-developed positive
WHOLE-ROCK MAJOR AND RARE EARTH
Eu-anomaly in association with the most prominent
ELEMENT GEOCHEMISTRY
HREE depletions observed in this data set, indicating
Data for whole-rock geochemistry for major and it is most likely the residue of a peraluminous melt
rare earth elements were obtained through Activa- that fractionated feldspar after it was extracted from
tion Laboratories, Ltd. in Ancaster, Ontario, a garnet-rich residuum (Figs S4e & S5).
Canada, with additional whole-rock major element
data acquired from the Ronald B. Gilmore XRF
PH A S E E Q U I L I B R I A M O D E L L I N G
Lab at the University of Massachusetts-Amherst
(Table S1). The data are combined with data from For the Opx-absent felsic granulite gneiss, Snoeyen-
Baldwin et al. (2006) to characterize the composi- bos et al. (1995) estimated UHT conditions ≥1000 °C
tional variation in the felsic granulites (Figs S4 & at ≥1.5 GPa based on ternary feldspar thermometry,
S5); in these figures, sample 04G-090A (Fig. 3b) rep- and assuming equilibrium with kyanite. The reliabil-
resents an example of the likely protolith, based on ity of this approach for estimating metamorphic
the previously mentioned field observations. The UHT conditions has been questioned by St ıpsk
a &
metagraywacke composition (CEVP) used in the Powell (2005) for similar felsic granulites in the Bohe-
partial melting experiments of Vielzeuf & Montel ıpska & Powell (2005) argued that the
mian Massif. St
(1994) and the compositions of the Bohemian Massif ternary feldspar was a relict of magmatic crystalliza-
felsic granulites were included in Fig. S4a,b for com- tion rather than metamorphic in origin. Baldwin
parison (from Fiala et al., 1987; Carswell & O’Brien, et al. (2006) adopted a similar interpretation for the
ıpsk
1993; St a & Powell, 2005; Tropper et al., 2005). Upper Deck felsic granulites, suggesting that the
Analysis of major element whole-rock geochemistry, felsic granulites were orthogneisses. However, the
following the approach of Frost et al. (2001), indi- REE data for the Upper Deck felsic granulite pro-
cates that the Upper Deck felsic granulite bulk com- tolith are consistent with a metasedimentary origin
positions are predominantly magnesian and (Fig. S4c). Furthermore, the identification of
peraluminous (Fig. S4a,b; Table S1). The protolith resorbed Ti-rich biotite inclusions in ternary feldspar
bulk composition (sample 04G-090A) is composi- from this study supports a metamorphic origin for
tionally similar to the greywacke bulk composition Upper Deck ternary feldspar, e.g. as a prograde
used in the experiments of Vielzeuf & Montel (1994; phase resulting from a fluid-absent biotite melting
Fig. S4a,b). Figure S4c illustrates the nearly one-to- reaction (Fig. 5b; Vielzeuf & Schmidt, 2001). The
one rare earth element normalization of the pro- most common way to establish UHT conditions is
tolith bulk composition relative to average bulk via high Mg–Al assemblages like sapphirine + quartz
compositions for Late Archean greywacke and (Goncalves et al., 2004; Kelsey & Hand, 2014;
Archean cratonic shale (Condie, 1993). Korhonen et al., 2014). Unfortunately, these assem-
The felsic granulite bulk compositions are quite blages have not been identified in the Upper Deck
variable, ranging from magnesian to ferroan and cal- domain, and most of the widespread felsic granulites
cic to alkalic–calcic (Fig. S4a,b; Table S1). This vari- have a mineralogy that limits what can be achieved
ability is attributed to accumulations of using conventional thermobarometric methods. In
garnet orthopyroxene (Fig. S4a,e) and feldspar this contribution, we use phase equilibria modelling
(Fig. S4b,d) in these residual bulk compositions. The with the bulk composition of the newly discovered
bulk compositions for some of the felsic granulites orthopyroxene-felsic granulite (sample 04G-097;
also share characteristics common to many Figs 7–10; Tables S1–S4) to provide new constraints
leucogranites, including the Opx-leucogranite dyke on this occurrence of UHT metamorphism.
derived from the felsic granulite that cuts mafic gran- The felsic granulite gneisses of the Upper Deck
ulite (sample 05G-051B in Fig. 3f; see leucogranite domain potentially experienced large degrees of par-
compositional field in fig. S4a,b from Frost et al., tial melting and melt loss, as suggested by abundant
2001). Some of the Opx-bearing felsic granulites have quartzofeldspathic leucosomes and garnet-rich layer-
low XAl bulk compositions (e.g. sample 04G-097; ing that define the early S1 foliation (Figs 3b–e & 4a,
XAl < 0.35; Table S1) similar to the Type I Opx-bear- b). These residual bulk compositions are suitable for
ing HP felsic granulites identified by Chen et al. determining the post-melt loss P–T conditions and
(2008). The entire suite of Upper Deck felsic gran- retrograde evolution of the felsic granulites. However,
ulites is generally more magnesian, alkaline and sil- constraining the prograde history of the felsic gran-
ica-poor than the Bohemian Massif felsic granulites ulites and understanding the significance of composi-
(Fig. S4a,b). Rare earth element data normalized to tional zoning in garnet and orthopyroxene requires
bulk continental crust values from Rudnick & Gao accounting for the effects of melt loss on the bulk
(2003) generally reveal heavy rare earth element composition. The inversion technique of Korhonen

04G-097: Opx-bearing felsic granulite

Na2O : CaO : K2O : FeO : MgO : Al2O3 : SiO2 : TiO2 : H2O
3.59 : 3.08 : 1.66 : 5.65 : 5.62 : 10.12 : 69.54 : 0.54 : 0.19 (mol%, initial bulk composition at stage 1)
3.58 : 2.97 : 1.74 : 5.52 : 5.39 : 9.98 : 69.18 : 0.51 : 1.14 (mol%, reintegrated bulk composition at stage 2)
3.54 : 2.59 : 1.95 : 4.98 : 4.65 : 9.54 : 68.42 : 0.43 : 3.91 (mol%, “protolith” bulk composition) + Qtz
1.8
+ Rt
Rt
Pl Grt
Opx Cpx L
Bt Kfs Grt Rt
Bt L
Bt L Kfs Bt Cpx
Fig. 7. Melt reintegration model Kfs Pl
L Kfs
Pl Grt L L Pl
l%
Bt Cpx Ms Bt Grt
represented as a composite pseudosection Pl
1 mo
L Pl Ms Grt
t Cpx
Pl Grt Rt Opx L Kfs
computed with Perple_X 6.6.8 for Opx- 1.6
Grt Grt Rt Grt Rt
elt =
Opx B
Rt
bearing felsic granulite 04G-097. Coloured Bt
contours correspond to mol.% melt m
vol m
Bt L Pl t Il
Pl tR
Grt Gr
isopleths. The far right panel corresponds Ms Opx
Opx L K fs
Bt L Kfs Bt xL
to the P–T pseudosection calculated for the Grt Op
6
Kfs Pl
Pl Grt Pl 1 Grt Rt
starting residual whole-rock bulk Bt L Grt
composition in Table S1 + 0.03 wt% H2O 1.4 Pl Grt
1
to achieve ~1 mol.% melt at the inferred P (GPa)
peak conditions. The four adjacent panels Opx L
Kfs Pl
illustrate the effect of melt reintegration Grt Rt
6
into the residual bulk composition such Ilm
that a melt connectivity threshold of 1.2 Opx Bt 2 Opx L Kfs

Bt
7 mol.% melt (~7 vol.%) was reached at Pl L Pl Grt Ilm
Grt
each stage. See text for discussion. Four Bt L Grt
melt reintegration stages (numbers 1–4) Pl Grt

Opx
Opx Bt
were performed at the 1 mol.% melt Bt
L Kfs
Pl
isopleth down pressure and temperature 1.0 Bt L Grt Pl Grt
Pl Grt Ilm
along the hypothetical P–T path depicted in
1
Fig. 8. The final reintegrated bulk Opx L Kfs
Pl Grt Ilm
composition at stage 4 is inferred to Opx
11
Bt
represent a plausible initial (protolith) bulk
6
6
L Kfs
3 Opx L Kfs Ilm
composition that was used to compute the Pl Grt
6
pseudosection in Fig. 8. The composition of 0.8 4

800 850 900 950 1000 1050
melt and the reintegrated bulk composition
for each stage are provided in Table S4. T (°C)
et al. (2013) was used to recalculate plausible pro- ing felsic granulite (05G-019B) bulk compositions
tolith bulk compositions by the stepwise reintegration was estimated assuming that all loss on ignition
of melt into the initial residual bulk composition reported in Table S1 corresponds to H2O. This
along an inferred prograde P–T path (cf. White amount of H2O was adjusted for the first stage of the
et al., 2004). In this context, the two main goals of melt reintegration by adding 0.03–0.04 wt% H2O to
the phase equilibria modelling are to explore the ensure that a minimum of 1 vol.% melt was present
effect of melt transfer on phase relations and con- at the inferred peak conditions, representing the
strain the P–T evolution for the felsic granulite amount of melt that would be retained as thin films
gneisses. Phase relations were then texturally linked along grain boundaries after melt loss (Yakymchuk
to monazite zoning and chemistry to provide & Brown, 2014a; see melt reintegration section
geochronological constraints on the evolution of fel- below). Although the amount of Fe2O3 is a critical
sic granulite gneisses in continental lower crust (see compositional variable in modelling Fe + Mg-bearing
monazite section below). partially molten systems, the impact on UHT
Pseudosections were computed in the (>950 °C) phase relations in the peraluminous pyr-
NCKFMASHTO and Na2O–CaO–K2O–FeO–MgO– rhotite-bearing (sample 04G-097) and graphite-bear-
Al2O3–SiO2–H2O–TiO2–Fe2O3 systems using Per- ing (sample 05G-019B) systems of this study appears
ple_X 6.6.8 (Connolly, 2005) and the updated 2002 to be limited. This effect was evaluated with
version of the internally consistent thermodynamic T XFe2 O3 ðXFe2 O3 ¼ Fe2 O3 =½FeO þ Fe2 O3 Þ pseudo-
data set of Holland & Powell (1998; hp02ver.dat in sections calculated at 1.45 GPa using the inferred
Perple_X; Figs 7–11 & S6). Phases and solution mod- protolith bulk compositions determined after melt
els considered in the modelling included: garnet reintegration for both samples (Fig. S6; Table S4).
(White et al., 2007), ternary feldspar (Fuhrman & The modelling indicates that virtually any addition of
Lindsley, 1988), orthopyroxene (White & Powell, Fe2O3 to the system will stabilize ilmenite (Fig. S6a).
2002), clinopyroxene (Green et al., 2007), biotite These results are contrary to petrographic observa-
(Tajcmanov a et al., 2009), ilmenite (White et al., tions from Opx-bearing felsic granulite gneisses that
2000), white mica (Coggon & Holland, 2002) and document the ubiquitous presence of rutile and the
haplogranitic melt (White et al., 2007; Table S3). The notable absence of ilmenite as inclusions in garnet
amount of H2O used in the modelling for the Opx- (Figs 5f & S1). Similar observations from the more
bearing felsic granulite (04G-097) and Sil + Ky-bear- typical Sil + Ky-bearing felsic granulite samples


3.54 : 2.59 : 1.95 : 4.98 : 4.66 : 9.54 : 68.42 : 0.43 : 3.91 (mol%, reintegrated bulk composition after stage 4) + Qtz
1.8 t
Bt Cpx Ms Bt Cpx R Bt Cpx L Opx Cpx L
rt
Grt Rt Pl Ms G rt Bt L Bt L Kfs Pl Kfs Pl Grt Kfs Grt
s G
Grt Rt M s Kfs Grt Rt Rt Rt
M
Pl PL Pl
L L
px x Ms
Bt Cpx Pl C Cp Opx Cpx L
1.6 Bt Bt
Grt Bt Cpx L Kfs Grt Rt
Kfs Pl Grt Rt
Ms Ilm Opx
Grt Pl Bt L Kfs Grt Rt L Kf
L rt Opx L Kfs Grt Rt
x
Cp m G Pl (PEAK ASSEMBLAGE FIELD)
Bt s Il Bt L Kfs Pl Grt Grt
M 1 Rt Melt loss event
1.4 modeled in Fig. 9
Bt L Pl
Ms Ilm Grt
Opx L K
fs Ilm ilm
Grt ou
Bt L Pl Ms Grt Rt t
P (GPa)
t
ou
s
1.2 2
M
ut
Bt o
Bt L PL Grt Opx L Kfs Ilm Grt
1.0
rt
Fig. 8. (a) P–T pseudosection derived with
H 2O-s
mG
Grt
3 Perple_X 6.6.8 using the reintegrated bulk
Pl Il
atura
lm
Bt lt
r me 7 composition for 04G-097 after the four-
Kfs
0.8 4
Pl I
Pl ath fo Fig.
P-T pration in Opx L Kfs Ilm
ted s
tL
stage melt reintegration procedure (Fig. 7;

Kfs
Grt g Opx Bt
reinte
xB
H2 O L Pl Grt
xL
Table S4). The black arrow corresponds to

olidu
Op
Op
Opx L Kfs Ilm

the inferred P–T path along which the melt
s
0.6 reintegration was performed. The numbers

700 750 800 850 900 950 1000 1050
(a) T (°C) denote the P–T conditions for each stage of
1.8 the modelled melt reintegration. Note the
0.1
8
Bt - TiO2 (wt%) 0
peak assemblage for sample 04G-097 at
0.1
4
Opx - Al203 (wt%) 3.
upper right: Opx + Grt + Rt + Ternary
5.5
2
0.1
Feldspar + Melt. Boundaries for Ms-out,
5.0
1.6
TiO
li
t 1 nu Bt-out, Ilm-out and the H2O-saturated
2
0 0
4.5
0.1 4.
in
0.1
6 Gr an
B
ch
solidus are highlighted with dashed lines.
t=
r i
s-
Gr
4.0
8
1.4 0.0 Location of the melt loss event modelled in
wt
0
5.
.%
Fig. 9 is denoted by the star. (b) High-T

3.5
P (GPa)
side of same P–T pseudosection in (a) with

onin
1.2 0.0
6
isopleths for mole fraction of grossular in
1-z
0
.% 7.
garnet. Note the highlighted fields
Grt
wt
3.0
6.0
x= corresponding to the measured
1.0 Op
in
0.04 es Al 2
O3
0
compositions of Grt1 cores and annuli and
Grs
= or 7.
Op
xc the P–T path between them inferred from
s ich grossular zoning. (c) Same P–T
0.8 Grt 1 core A l-r
or 6.
0
Grs
-po pseudosection as in (b) with isopleths for
Grt - Grossular wt% TiO2 in biotite and wt% Al2O3 in
0.6 orthopyroxene. Note the highlighted field
800 850 900 950 1000 1050 850 900 950 1000 1050
corresponding to core compositions
(b) T (°C) (c) T (°C) measured in Opx for 04G-097.
support reducing conditions (XFe2 O3 0.05; The first stage of melt integration was performed in
Fig. S6b), including the absence of ilmenite and the the stability field where the peak metamorphic assem-
presence of graphite as inclusions in garnet and blage (Opx + L + Kfs Pl + Grt + Rt + Qtz) was
throughout the matrix (Fig. S2). observed at >1.4 GPa and >925 °C (Fig. 7). The pre-
cise P–T location of the melt integration step is not
critical as the melt composition does not vary
Orthopyroxene-bearing felsic granulite 04G-097: melt
significantly within this stability field (cf. Korhonen
reintegration and constraints on prograde evolution and
et al., 2013). We selected a point located on the
peak metamorphism
1 mol.% melt isopleth at a pressure of 1.45 GPa
The Opx-bearing felsic granulite (04G-097) preserves (Fig. 7), corresponding to the amount of melt that
evidence for a prograde evolution involving the pro- would be retained as thin films along grain boundaries
duction of Al-rich Opx and grossular-poor garnet after melt loss (Yakymchuk & Brown, 2014a). At this
cores that transition into grossular-rich garnet annuli point, 6 mol.% melt was added to the bulk composi-
associated with Ti-rich biotite inclusions and poly- tion to reach a value of 7 mol.% melt in the rock. This
phase inclusions of Kfs + Pl + Qtz + Bt + Rt. volume of melt is analogous to the maximum thresh-
Because this rock was affected by melt loss, the old for melt interconnection and transfer based on
down-temperature melt reintegration approach of experimentally constrained observations of the melt
Korhonen et al. (2013) was applied to reconstruct a connectivity transition (Rosenberg & Handy, 2005).
plausible protolith bulk composition to model the The new bulk composition obtained after melt reinte-
prograde phase relations. gration was used to compute a new pseudosection


3.54 : 2.59 : 1.95 : 4.98 : 4.66 : 9.54 : 68.42 : 0.43 : 3.91 (mol%, reintegrated bulk composition) X = 0
3.58 : 3.01 : 1.63 : 5.46 : 5.55 : 10.06 : 69.82 : 0.55 : 0.36 (mol%, after 100% melt loss) X = –100%
3.53 : 2.49 : 2.03 : 4.86 : 4.44 : 9.42 : 68.08 : 0.40 : 4.76 (mol%, after 33% melt gain) X = +33% + Qtz
Opx L
Kfs Grt - Grossular
Ilm
Grt Opx L Kfs 0.08
1050 Ilm Grt Rt
t
0.09
ou
Opx L Kfs Grt Rt

Il m
0.10
Opx Bt Kfs Grt Rt
1000 Opx Bt Kfs Grt Rt Ilm
Opx Bt Kfs Grt Ilm 0.11
Opx L Kf Pl Grt Rt 0.12

Bt out
T (°C)
950 0.13
Opx Opx out

0.14
Me
Bt
lt in
900 Kfs Bt Bt L Kfs Grt

Pl Kfs
Grt Grt
Rt Bt
850 px Kfs 0.15
+C Pl Bt L Kfs Pl Grt
Bt Kfs
Grt Pl Grt
Rt (b)
(a)
–66% –33% 0 +33%
Opx - Al203 (wt%) X 04G-097
4.5
melt loss melt gain
4.0
1050
3.5
1000
3.0
T (°C)
950 2.5
Fig. 9. (a) T–X pseudosection computed at 1.45 GPa
with Perple_X 6.6.8 for Opx-bearing felsic granulite
04G-097 using the melt reintegrated bulk composition
900 in Fig. 8. The X-axis corresponds to this bulk
composition (at X = 0) affected by melt transfer,
ranging from 100% melt loss to 33% melt gain. The
2.0 melt composition and volume at X = 0 were determined
850 at 1.45 GPa and 965 °C using the pseudosection in
Fig. 8. Boundaries for Ilm-out, Bt-out, Opx-out and
Melt-in are highlighted with dashed white lines. Black
and white arrows indicate T–X paths discussed in the
(c) text. (b) Same T–X phase diagram as in (a) with
–66% –33% 0 +33% isopleths for mole fraction of grossular in garnet. (c)
X 04G-097 Same T–X phase diagram as in (a) with isopleths for wt
melt loss melt gain % Al2O3 in orthopyroxene.
(Fig. 7; Table S4). Down-temperature melt reintegra- reintegration steps were defined where the P–T path
tion was carried out along an inferred prograde P–T crossed the 1 mol.% melt isopleth (Fig. 7).
path corresponding to the burial of sediments to a Four melt reintegration steps, representing addition
depth in excess of 50 km (>1.4 GPa; P–T path in of ~25 mol.% melt, were necessary to produce a
Figs 7 & 8). The low-P portion of the path represents plausible protolith bulk composition such that the
burial of the sediments from the surface to 850 °C and solidus was H2O-saturated (Fig. 8). The primary
0.85 GPa. For >850 °C, it is assumed that the temper- effect of adding melt to the initial residual bulk com-
ature during burial to > 1.4 GPa was buffered by the position was the addition of H2O (Table S4). The
fluid-absent biotite melting reaction. Subsequent melt main features of the pseudosection computed with

04G-097: residue = reintegrated bulk composition - 90% melt at 1.45 GPa and 965°C
3.58 : 2.96 : 1.67 : 5.40 : 5.44 : 9.99 : 69.64 : 0.53 : 0.80 (mol%) + Qtz
1.8
Bt Cpx Kfs Ms
Grt Ky Rt Bt Kfs Pl Ms Grt Ky Rt
Bt Cpx Rt Bt Cpx Kfs Pl Bt Cpx Kfs
Kfs Grt K y Ms Grt Ky Rt
rt Pl Grt Rt
1.6
Ky Rt
Pl
G Rt
s y Bt Kfs Pl Grt Rt
Kf tr K 0.
16
x G px
Cp Pl tC t
xB tR
Bt Kfs Op Pl Gr
Bt Kfs Kfs Grs-rich Grt2
Bt Cpx rt Rt corona +
G
1.4 Pl Ilm x Bt Kfs Pl Al-poor Opx rims
fs OpIlm Grt Rt
px K
P (GPa)
rt R t B t C rt
G
Pl Ilm Opx l Ilm G Opx Bt
Bt Kfs P Kfs Pl wt
%
.5
lm Grt Bt Cpx Kfs Opx Bt Cpx Kfs Grt Rt = 2
l I Pl Ilm Grt Rt x
1.2 fs P Pl Ilm Grt 2.0 Op
Bt K 1.5 0.1
4 in
0.1
6
rt = O3 3.0
inG Al 2
Grs
3.5
Opx Bt Kfs Pl Ilm Grt
1.0 4.0
2
0.1
Opx Bt L Kfs 0 8
Pl Ilm Grt 0.1 0.0 0.06
0.8
650 700 750 800 850 650 700 750 800 850
(a) (b)
T (°C) T (°C)
Fig. 10. (a) P–T pseudosection derived with Perple_X 6.6.8 for residual Opx-bearing felsic granulite bulk composition for sample
04G-097 (in Fig. 8) after 90% melt loss at 1.45 GPa and 965 °C. (b) Same P–T phase diagram as in (a) with isopleths for mole
fraction of grossular in garnet and wt% Al2O3 in orthopyroxene. Note highlighted melt-absent field defining compositions for
high-grossular Grt2 and recrystallized low-Al Opx measured in 04G-097.
the derived protolith composition are consistent with Pl + Qtz + Kfs + Rt (Fig. 5f), interpreted as relicts of
constraints from partial melting experiments using a crystallized melt. The garnet zoning and textures
greywacke bulk composition (Fig. 8; Vielzeuf & Sch- are most consistent with a loading path from 0.6–
midt, 2001). These features include: (i) a water-satu- 0.8 to >1.4 GPa. This path was likely buffered in
rated solidus at 650–675 °C, (ii) the fluid-absent temperature by the low-variance fluid-absent biotite
muscovite melting reaction at >1.0 GPa between 700 melting reaction and crossed the grossular isopleths
and 800 °C (Ms-out in Fig. 8a), (iii) a large P–T at a relatively high angle before reaching the peak
region between 700 and 875–950 °C where the incon- assemblage stability field of Opx + L + Kfs + Grt +
gruent biotite melting reaction produces peritectic Rt at UHT (Fig. 8a,b). The inferred loading path
garnet + melt and (iv) a steep and narrow low- in Fig. 8b coincides with increasing vol.% garnet,
variance field between 875 and 950 °C corresponding modest amounts of melt production (~25 vol.%)
to the fluid-absent biotite melting reaction and the and a diminishing abundance of ilmenite, consistent
production of peritectic Opx + Kfs + Grt (Fig. 8a). with petrographic observations. Measured and mod-
The high-T side of the pseudosection derived from elled Ti contents up to 6.25 wt% TiO2 in resorbed
the melt reintegrated bulk composition was con- biotite inclusions in garnet and ternary feldspar are
toured for mole fraction of grossular in garnet consistent with prograde partial melting of biotite
(Fig. 8b), wt% TiO2 in biotite and wt% Al2O3 in at temperatures in excess of 900 °C (Fig. 8c). The
orthopyroxene (Fig. 8c). Grossular-poor garnet cores presence of rutile is critical for constraining a
(Grs3–4) reported for Grt1 are most compatible with minimum pressure of ~1.3 GPa at >950 °C (above
low-P growth at 0.6–0.8 GPa and 800–900 °C ‘Ilm-out’ in Fig. 8a). Peak metamorphic conditions
(Fig. 8b). This garnet composition would be in equi- were refined by using the maximum measured grossu-
librium with aluminous orthopyroxene (4.5–6 wt% lar content of the Grt1 annuli (up to 12 1%) in
Al2O3; Fig. 8c), consistent with the core compositions equilibrium with orthopyroxene and rutile, corre-
measured in orthopyroxene porphyroblasts (up to sponding to 1.35–1.8 GPa and 950–1100 °C (Fig. 8b,
4.48 wt% Al2O3; Table S2). The low-grossular cores c). At these peak conditions, the Al2O3 content of
of Grt1 porphyroblasts are surrounded by grossular- orthopyroxene would decrease to ~3 wt% Al2O3
rich annuli (up to Grs13) that contain Ti-rich biotite (Fig. 8c), consistent with the observed core-to-rim
(6.07 wt% TiO2) and poly-phase inclusions of Bt + zoning (Fig. 5c).

05G-019B: Sil + Ky-bearing felsic granulite

1.54 : 1.56 : 2.11 : 10.30 : 5.37 : 12.24 : 66.05 : 0.66 : 0.16 (mol%, initial bulk composition at stage 1)
1.62 : 1.55 : 2.19 : 9.79 : 5.09 : 12.08 : 66.07 : 0.62 : 0.97 (mol%, reintegrated composition at stage 2)
1.85 : 1.38 : 2.37 : 8.30 : 4.30 : 11.39 : 65.67 : 0.52 : 4.23 (mol%, “protolith” composition) + Qtz
1.6
Bt Bt L
Kfs Kfs L Kfs Grt
07 Ky Rt
Pl Ms Pl Ms 0.
Grt Grt Ky
1
Rt Rt
0.07
L Kfs Pl
08
0. Grt Ky Rt
05
06 1
0.
Bt Kfs Pl 0.
Ms Grt t=
Ky Rt Gr
in
1.4 s
Fig. 11. (a) Melt reintegration model Gr
represented as a composite pseudosection Bt L Kfs Pl
computed with Perple_X 6.6.8 for Sil + Ky- Bt Kfs
Grt Ky Rt L Kfs Grt
Sil Rt
bearing felsic granulite 05G-019B. Coloured Pl Ms
7
0.0
Grt Ky ite
ol %
contours correspond to mol.% melt an ite
Ky an
isopleths. Dashed black lines represent m
6m
07
Bt li li
0.
S
isopleths for mole fraction of grossular in
P (GPa)
Pl
lt =
Ms
04
garnet. The far right panel corresponds to
me
1.2
0.
Grt Bt L 5
the P–T pseudosection calculated for the Ky 0.0
6
Kfs
vol
starting residual whole-rock bulk Pl Ms L Kfs Pl
Grt Ky
composition in Table S1 + 0.04 wt% H2O Bt
Grt Sil Rt
to achieve ~1 mol.% melt at the inferred L

Pl
peak conditions. The four adjacent panels Ms Bt L Kfs Pl
Bt L Kfs
0.0
4 L Kfs
Grt Pl Grt Rt Sil L Kfs Grt Sil
illustrate the effect of melt reintegration Ky
Grt Ky Pl Grt Rt Ilm
into the residual bulk composition such Sil Rt
Ilm
that a melt connectivity threshold of 1.0 2
6
1
7 mol.% melt (~7 vol.%) was reached at

each stage. Four melt reintegration stages
1
(numbers 1–4) were performed at the L Kfs Pl 3

1 mol.% melt isopleth down pressure and Bt L Kfs Pl Grt Ilm Sil 0.0
Grt Sil Bt L Kfs
temperature along a hypothetical P–T path Pl Grt Ilm Sil
similar to the path imposed for sample 3 3
0.0
11
04G-097 in Figs 7 and 8. The composition
1
of melt and the reintegrated bulk 0.8 4

800 850 900 950 1000
composition for each stage are provided in
Table S4. T (°C)
We infer that partial melting and the formation of Figure 9 is a T–X pseudosection calculated at
Opx + Grt1 occurred during a loading path from 1.45 GPa. The X-axis corresponds to the melt reinte-
0.6–0.8 to >1.35 GPa at a temperature that was buf- grated bulk composition of 04G-097 affected by a
fered by the fluid-absent biotite melting reaction at range of melt transfer from 100% melt loss to 33%
~900–950 °C. The pseudosection in Fig. 8 that facili- melt gain. The starting melt composition and volume
tates this inference is based on the assumption that at X = 0 were determined at 1.45 GPa and 965 °C
the entire prograde evolution occurred under a closed using the pseudosection in Fig. 8. These P–T condi-
system. In the following section, we explore the effect tions are located just above the completion of the
of melt transfer at peak metamorphic conditions. fluid-absent biotite melting reaction (star in Fig. 8a).
At peak metamorphic conditions above 950 °C, melt
transfer has no significant effect on phase relations
Orthopyroxene-bearing felsic granulite 04G-097: effect of
(Opx + L + Kfs + Grt + Qtz field in Fig. 9a) or min-
melt transfer
eral composition (Fig. 9b,c). The most prominent
The development of grossular-rich (up to Grs16) effect of melt loss is visible at temperatures lower
coronitic Grt2 around low-Al dynamically recrystal- than 950 °C. If isobaric cooling occurred under
lized orthopyroxene (1.08 wt% Al2O3) cannot be closed system conditions (white arrow in Fig. 9a),
explained by the pseudosection in Fig. 8. Although orthopyroxene would become unstable in the pres-
the Grs16 isopleth in Fig. 8c occurs at P–T conditions ence of melt. In contrast, significant melt loss (90%
of >1.4 GPa and 800–925 °C, it is noteworthy that or more) followed by an episode of isobaric cooling
this garnet composition would be in equilibrium with (black arrows in Fig. 9a) would stabilize low-Al
a melt-bearing assemblage that does not contain orthopyroxene at sub-solidus conditions (Fig. 9c).
stable orthopyroxene. To reconcile observations with Figure 10 corresponds to a P–T pseudosection cal-
the predicted phase relations, we carried out several culated with the residual bulk composition obtained
models to constrain the effect of melt transfer. after a loss of 90% of the melt produced at 1.45 GPa

and 965 °C. The main effect of melt loss is to stabi- annuli in equilibrium with L + Kfs + Ky + Rt would
lize low-Al orthopyroxene in equilibrium with high- have been stable at >1.4 GPa and >975 °C (Fig. 11),
grossular garnet at sub-solidus conditions (Figs 9a & consistent with the original conclusions of Snoeyen-
10). Another notable effect of melt loss is the bos et al. (1995).
stabilization of ilmenite to lower temperatures (cf. The second-generation low-phosphorous garnet
Figs 8a & 10a). Although rutile is the main Ti-bear- rim (Grt2 in Fig. 6) is characterized by a low-grossu-
ing phase at peak metamorphic conditions based on lar composition (Grs3–5) intergrown with retrograde
the abundant occurrence of rutile inclusions in Grt1 plagioclase (Fig. 6a,b). The stability of low-grossular
(Fig. 5f), ilmenite is an abundant matrix phase that garnet + plagioclase in the absence of melt is largely
has been observed in contact with rutile in the vicin- outside the range of P–T conditions modelled in
ity of coronitic Grt2 which, in turn, surrounds Fig. 11. We suggest that Grt2 may have grown in
orthopyroxene and Grt1 (Fig. 5a). Mineral composi- equilibrium with plagioclase as a sub-solidus assem-
tion contours suggest maximum P–T conditions for blage during a period of reheating after cooling to an
the formation of low-Al orthopyroxene and coronitic ambient crustal geotherm. A different effective bulk
Grt2 in association with ilmenite at 600–675 °C and composition would be required to accurately model
1.0–1.2 GPa (highlighted field in Fig. 10b). the conditions for growth of Grt2.
Sillimanite + kyanite-bearing felsic granulite 05G-019B: MONAZITE TRACE ELEMENT ANALYSIS AND
P–T conditions GEOCHRONOLOGY BY EPMA
Precise thermobarometric constraints from the more The procedure for Th–U–total Pb monazite
widespread aluminosilicate-bearing felsic granulite geochronology and trace element analysis by EPMA
gneisses in the Upper Deck domain are limited by in this study follows the work of Jercinovic & Wil-
their high-variance mineral assemblages (Sil Ky + liams (2005), Williams et al. (2006, 2007), Jercinovic
Kfs + Grt + Rt + Qtz). Textural observations support et al. (2008) and Dumond et al. (2008). Calibrated
equilibrium of ternary feldspar with kyanite, an infer- overlap correction factors for peak interferences of
ence that requires minimum P–T conditions of YLc on PbMa, ThMf1 and ThMf2 on PbMa,
~1.5 GPa and 950 °C (Snoeyenbos et al., 1995). We second-order LaLa on PbMa, ThMc on UMb, KKa
explored the plausibility of these conditions along a on Umb, and NdLb3 on EuLa were applied prior to
prograde loading P–T path similar to that inferred ZAF corrections during the analytical sessions (see
previously and modelled for the orthopyroxene-bear- Donovan et al., 1993; Pyle et al., 2002, 2005; Jerci-
ing felsic granulite (Fig. 11). The residual bulk com- novic & Williams, 2005; Jercinovic et al., 2008). Full
position for 05G-019B was used for the first stage of thin-section X-ray maps were collected via EPMA to
the melt reintegration (Fig. 11; Table S4). The initial identify all monazite grains (e.g. Fig. S3; following
H2O content was adjusted (0.04 wt% 0.16 mol.%) Williams & Jercinovic, 2012). Monazite grains were
to ensure that 1 mol.% melt was present at the mapped at high spatial resolution (0.3–0.5 lm step
inferred peak conditions, representing what would sizes) at 15 kV and 200 nA for 70–100 ms per pixel
have been melt films retained after melt loss. Subse- on a CamecaÒ SX50 electron microprobe. X-ray
quent melt reintegration steps along the P–T path in maps for YLa, CaKa, ThMa and UMb were pro-
Fig. 11 involved the addition of 6 mol.% melt to the cessed both simultaneously and then, individually to
bulk composition and recalculation of a new pseudo- identify similar and compositionally distinct domains
section, as discussed previously for sample 04G-097. to guide subsequent quantitative analysis. Details
Each pseudosection was contoured with isopleths for regarding the analytical methods for monazite trace
mol.% melt and mole fraction of grossular in garnet element analysis by EPMA and the determination of
(Fig. 11). each domain-specific date and error are summarized
The composite P–T pseudosection modelled with in Williams et al. (2006) and Dumond et al. (2008).
the kyanite-bearing felsic granulite bulk composition All dates are plotted as histograms scaled relative to
reveals two large stability fields at >1.0 GPa and a consistency standard (Williams et al., 2006). The
>850 °C where melt is in equilibrium with Kfs Pl + standard used in this study is the Moacir Brazilian
Grt + Ky (or Sil) + Rt (Fig. 11). The low-T and low- pegmatite monazite with a weighted mean 207Pb/235U
P boundaries of these stability fields correspond to age of 504.3 0.2 Ma (2r, MSWD = 0.64) (Gasquet
the appearance of biotite and ilmenite respectively et al., 2010).
(Fig. 11). The continuous change in high-phospho- Twenty-nine compositional domains were analysed
rous garnet composition (Grt1) from a low-grossular in monazite grains in samples 04G-097, 05G-019B and
core (Grs3-4) to a high-grossular annulus (Grs7-11) 05G-026B using the CamecaÒ SX100 Ultrachron elec-
and the garnet core-to-rim transition from sillimanite tron microprobe at the University of Massachusetts-
inclusions to kyanite inclusions (Fig. 6a) are consis- Amherst. The span of ages from 2.66 to 1.89 Ga is
tent with the inferred loading path illustrated in similar to that identified by Baldwin et al. (2006). The
Fig. 11. The peak assemblage of high-grossular Grt1 data set consists of matrix grains, inclusions in garnet

and intergrowths with garnet rims (Table S5). Textu- monazite cores and overgrowths range from
ral observations were combined with major and trace 2613 18 to 2552 19 Ma (2r) with a marked
element data for each monazite domain to link mon- enrichment in Th (4.79–20.00 wt% ThO2) and
azite to episodes of partial melting, garnet growth and depletion in Y (<800 ppm) relative to domain 1
feldspar recrystallization (see Discussion). (Figs 12a & 13c,d; Table S5). Domain 3 low-Th mon-
azite rims (0.05–6.20 wt% ThO2) tend to yield dates
with large uncertainties ranging from 2664 247
Monazite compositional zoning and textural setting
to 2514 47 Ma (2r; Figs 12a,b, 13d,e & 14b;
Four distinct compositional domains are present in Table S5). Domain 3 monazite rims are also distinctly
monazite grains from all felsic granulite samples. depleted in Ca (0.07–0.91 wt% CaO) and Sm (below
Domain 1 monazite consists of rare resorbed high-Y detection – 0.92 wt% Sm2O3) with respect to domains
cores (Figs 12a & 13a–c). Domain 2 monazite, which 2 and 3 (Table S5). Four of the analysed grains with
consistently occurs as overgrowths around domain 1 domain 3 rims are completely included in the high-
cores and locally contains inclusions of K-feldspar grossular annulus of Grt1 (Figs 5e, 12a,b & 13e; grains
(Figs 12a & 13c), is defined by high-Th, high-Ca and 05G-019B-YZ-1-m18 and 05G-019B-YZ-2-m9 in
low-Y (Figs 12a & 13c) and includes cores of grains Table S5). Comparison of domain 2 core and domain
with oscillatory or sector zoning defined by Th 3 rim data for three of the monazite grains reveals
(Fig. 13d). Domain 3 monazite is defined by low-Th, marked enrichments in XLaPO4 þ XCePO4 and XEuPO4
low-Ca rims on domain 2 monazite (Figs 12a,b & from core to rim (arrows in Fig. S8a,b). The four
13d,e). Monazite domains 2 and 3 have each been domain 3 monazite rims in sample 05G-019B have
observed at least once in the high-grossular annulus some of the largest Eu-anomalies in the entire data set
of Grt1 (Figs 12b & 13e). Domain 4 monazite pre- (7.6–14.4) and together yield a weighted mean age of
dominantly occurs as thin rims (10 lm-wide or less) 2539 21 Ma (2r, MSWD = 3) for growth of Eu-
that are slightly enriched in Th + Ca relative to enriched, Th- and Ca-depleted monazite (Fig. S8c;
domain 3 monazite (Figs 12c,d & 13f). Less com- Table S5). The relatively large MSWD may indicate a
monly, domain 4 occurs as rims on monazite inclu- period of protracted monazite growth, instead of a sin-
sions in the fractured rims of low-phosphorous Grt2 gle discrete event. Domain 4 monazite grains are
in sample 05G-019B (Fig. 13f) and as intergrowths highly depleted in Th (0.07–1.41 wt% ThO2) and are
with Grt2 (Fig. 13g). More commonly, domain 4 generally the most enriched in La + Ce (Fig. S8a;
monazite occurs as outermost rims on matrix grains Table S5). Dates for domain 4 monazite range between
(Figs 12c,d, 13h & 14b) which are associated with 2020 355 and 1885 20 Ma (2r; Figs 12c,d, 13f–h
recrystallized ribbons and grains of ternary feldspar & 14b; Table S5). Five dates for domain 4 matrix
(Figs 12d, 13h & 14a). Thorium zoning in domain 4 monazite grains and monazite cores or rims included
rims of matrix monazite locally defines asymmetric in or intergrown with Grt2 rims (Fig. 13f,g) yield a
overgrowths consistent with dextral, top-to-the-NE weighted mean age of 1913 25 Ma (2r,
oblique-slip displacement in Opx-bearing felsic gran- MSWD = 0.1) for growth of domain 4 monazite
ulite (sample 04G-097; Fig. 12d) and elongate syn- (Fig. S8c). Three dates for elongated overgrowths of
tectonic overgrowths in D2 dextral ultramylonite domain 4 monazite in the 05G-026B ultramylonite
zones (sample 05G-026B; Figs 14b & S3). provide a weighted mean age of 1924 72 Ma (2r,
MSWD = 0.2) for growth of domain 4 monazite
(Figs 14b & S8c; Table S5). Three dates for syn-kine-
Monazite geochemistry and geochronology
matic dextral overgrowths in the 04G-097 Opx-felsic
Geochemical data and calculated Th–U–total Pb dates granulite yield a weighted mean age of 1897 11 Ma
are summarized in Figs S7 and S8 and Table S5. Many (2r, MSWD = 2) for dextral shearing and growth of
of the dates yield 2r uncertainties in the order of domain 4 monazite (Figs 12c,d & S8c; Table S5).
18–55 Ma. Large errors (in excess of 60 Ma) for
some monazite domains discussed below are attributed
DISCUSSION
to low-Th concentrations (≤0.78 wt% ThO2 in
Table S5). Most domains display a positive Eu-anom- Phase equilibria modelling and melt reintegration
aly (Eu* = 1.5–18.9) in contrast to nearly all published techniques applied to residual peraluminous bulk
data which depict negative Eu-anomalies in monazite compositions is consistent with fluid-absent biotite
(Fig. S7; Table S5; see Discussion; Bea, 1996; Bea & melting and peritectic garnet growth along a pro-
Montero, 1999; Hermann & Rubatto, 2003; Kelly grade loading path from ≤0.8 to ≥1.4 GPa, culminat-
et al., 2006; Rubatto et al., 2006). All monazite ing with UHT conditions (>950 °C; Figs 8 & 11).
domains, with the exception of domain 1, are low in Y These results are in agreement with previous P–T
(<800 ppm Y; Table S5). estimates determined for the felsic granulite gneisses
The high-Y (1.46–1.84 wt% Y2O3) domain 1 mon- (Snoeyenbos et al., 1995) and for mafic granulite
azite cores were dated at 2636 20 to 2615 21 Ma gneisses interlayered with the felsic gneisses (Snoeyen-
(2r; Figs 12a & 13a,b; Table S5). Dates for domain 2 bos et al., 1995; Baldwin et al., 2003). Melt must

(a) 04G-097-2-m1 [inclusion in Grt1] Grt1

Domain 1 core
1.84 wt.% Y
4.53 wt.% ThO2
2636 +/– 20 Ma
Kfs Kfs
Domain 2 rim
Domain 3 outer rim
641 ppm Y
206 ppm Y
13.52 wt.% ThO2
6.20 wt.% ThO2
1.42 wt.% CaO
0.91 wt.% CaO
2552 +/– 19 Ma
2527 +/– 28 Ma
Y Lα Th Mα Ca Kα
50 µm = Background acquisition point = EPMA analytical point
(b) (c) Th Mα
Bt
Rt
Bt
Grt1 Grt1 04G-097-2-m7

matrix monazite
Domain 3 rim
123 ppm Y Domain 4 rim
Rt 0.14 wt.% ThO2 1885 +/– 20 Ma
0.14 wt.% CaO 6.57 wt.% ThO2
2595 +/– 219 Ma
04G-097-2-m5
BSE Ca Kα [inclusion in Grt1] Y Lα
200 µm 50 µm 50 µm
SW NE
04G-097-1-m1
(d) Th Mα Domain 4 rim
1890 +/– 20 Ma
4.80 wt.% ThO2
Fig. 5b
Y Lα
matrix monazite
50 µm Domain 4 rim
1911 +/– 17 Ma
2 mm 4.63 wt.% ThO2
Fig. 12. Monazite maps and data for Opx-bearing sample 04G-097. Grains in (a) and (b) are inclusions in the high-grossular
annulus of Grt1. Grains in (c) and (d) occur in the matrix. (a) X-ray maps of monazite grain 2 m1 with EPMA Th–U–total Pb
dates (2r). (b) Backscattered electron (BSE) image of textural setting of monazite grain 2-m5 in garnet with X-ray map depicting
EPMA Th–U–total Pb date (2r). (c) X-ray maps of monazite grain 2-m7 with EPMA Th–U–total Pb date (2r). (d) Plane
polarized light photomicrograph illustrating setting of monazite grain 1-m1 beneath a large dextral ternary feldspar r-clast. Note
syn-kinematic rims on 1-m1 depicted in the ThMa map and EPMA Th–U–total Pb dates (2r) on rims in the YLa map.
have been lost from these residual felsic granulites to vide a basis for inferring monazite petrogenesis with
facilitate preservation of peak metamorphic assem- respect to prograde and retrograde silicate reactions
blages (Brown, 2002; White & Powell, 2002; White (see review by Williams et al., 2007). What remains less
et al., 2004). These results provide context for well-understood, however, is the behaviour of Y, Th
understanding the petrogenesis of monazite in conti- and other elements in monazite during partial melting
nental lower crust. or melt extraction (Watt & Harley, 1993; Watt, 1995;
Compositional zoning defined by X-ray maps, in Kohn et al., 2005). Establishing links between
combination with major and trace element data, pro- growth/dissolution/re-precipitation of monazite and

(a) (b) 05G-019B-YZ-2-m12 (c) 05G-019B-XZ-2-m11

05G-019B-XZ-2-m16
[inclusion in Grt1] [Inclusion in Grt1] [Inclusion in Grt1]
Domain 1 core
1.56 wt.% Y Domain 1 core
2616 +/– 28 Ma 1.46 wt.% Y
2615 +/– 21 Ma
Kfs
Domain 2 rim
798 ppm Y
Y Lα 10 µm Y Lα 10 µm Y Lα 2575 +/– 20 Ma
10 µm
(d) (e) Domain 3 rim

05G-019B-XZ-m6 Ca Kα 2514 +/– 47 Ma
[Inclusion in Grt1
1.76 wt.% ThO2
- see Fig. 6a]
0.21 wt.% CaO
Domain 3 rim 4396 ppm Eu
2587 +/– 39 Ma
1.85 wt.% ThO2 05G-019B-YZ-2-m9
Domain 2 core
0.48 wt.% CaO [Inclusion in Grt1]
2605 +/– 21 Ma
4660 ppm Eu
314 ppm Y
9.34 wt.% ThO2
1.63 wt.% CaO 10 µm
3905 ppm Eu
Grt1
Th Mα 10 µm
200 µm
Th Mα
(f) 05G-019B-XZ-2-m24 (g) (h) 20 μm
[Inclusion in Grt2
BSE U Mβ
- see Fig. 6b] 05G-019B Kfs
-XZ-2-m10
[Grt2 + Mnz Domain 4 rim Pl
intergrowth] 0.10 wt.% ThO2
1894 +/– 164 Ma
Qtz Kfs
Domain 4 rim Grt2 Domain 4 core
1906 +/– 55 Ma 1907 +/– 48 Ma
1.41 wt.% ThO2 1.39 wt.% ThO2
Th Mα Ca Kα 20 µm 20 µm 05G-019B-YZ-1-m17
[matrix monazite]
Fig. 13. Monazite maps and data for Ky + Sil-bearing sample 05G-019B. (a–d) X-ray maps of monazite inclusions in garnet with
EPMA Th–U–total Pb dates (2r). (e) Ca Ka map of garnet depicting location of monazite grain YZ-2-m9 in high-grossular
annulus of Grt1 with EPMA Th–U–total Pb date (2r). (f) X-ray maps of monazite grain XZ-m24 included in Grt2 rim with
EPMA Th–U–total Pb date (2r). (g) Backscattered electron image of monazite intergrowth with Grt2 rim with location of
EPMA Th–U–total Pb date (2r). (h) X-ray map of matrix monazite grain with EPMA Th–U–total Pb date (2r).
the appearance or loss of melt or a major silicate phase dance in crustal granulites are used to constrain a
requires knowledge of REE-residence in the various model for the role of monazite as a monitor of partial
phases (Bea, 1996; Jung & Hellebrand, 2006). Our melting, garnet growth and feldspar recrystallization
results combined with existing data on REE abun- in continental lower crust (Fig. 15).

2 monazite cores and overgrowths are depleted in Y

(a) BSE 100 µm and enriched in Th and Ca relative to domain 1. Mon-
azite and garnet are the only Y-bearing phases in the
Qtz felsic granulite gneisses, and thus the depletion of Y
Qtz observed in domain 2 implies growth of monazite dur-
05G-026B-m1 ing or following the appearance of garnet. The occur-
Ap rence of Y-depleted domain 2 monazite as inclusions
Qtz Mnz in Grt1 in samples 04G-097 and 05G-019B supports a
syn-garnet interpretation for monazite growth between
2.61 and 2.55 Ga (stage 3 in Fig. 15). Ca-enrichment
in domain 2 monazite may be due to breakdown of
Qtz
plagioclase during melting and/or dissolution of apa-
tite.
Kfs The dramatic enrichment in Th is interpreted in
the context of two fluid-absent melting reactions
(Fig. 15). For the orthopyroxene-bearing felsic gran-
(b) SW NE ulite, the results support a reaction of the type Bt +
Pl + Qtz + high-Y, low-Th Mnz ? Grt + Opx + Kfs
Th Mα Domain 3 2596 +/– 54 Ma 0.70 wt.% ThO
2
+ Melt + low-Y, high-Th Mnz. For the aluminosili-
core cate-bearing felsic granulite, the results support a
reaction of the type Bt + Pl + Sil + Qtz + high-Y,
low-Th Mnz ? Grt + Kfs + Melt + low-Y, high-Th
Mnz. Preferential incorporation of Th and loss of Y
in domain 2 monazite is most consistent with growth
of monazite in the presence of melt + garnet
Domain 4 Domain 4 K-feldspar. This is supported by the occurrence of
1926 +/– 117 Ma 1916 +/– 99 Ma
K-feldspar inclusions (inferred to represent products
0.39 wt.% ThO2 10 μm 0.40 wt.% ThO2
of the melting reaction) that cut across domain 1 and
= Background acquisition point are overgrown by domain 2 monazite in both types
= EPMA analytical point of felsic granulite (Figs 12a & 13c).
Incorporation of Th into monazite in the presence
Fig. 14. Monazite maps and data for felsic granulite of melt requires the partition coefficient for Th
ultramylonite sample 05G-026B. (a) High-contrast between monazite and melt to be greater than 1
backscattered electron (BSE) image of textural setting of (Rapp et al., 1987), as recently demonstrated experi-
elongate monazite in the large recrystallized quartz ribbon
illustrated in Fig. S3. (b) ThMa map of same monazite grain mentally by Xing et al. (2013) for monazite in haplo-
with EPMA Th–U–total Pb dates (2r). granite melts. Evidence for compatible behaviour
between Th and monazite is implied by high-Th oscil-
latory-zoned grains that arguably crystallized in the
Monazite as a monitor of garnet growth and melting
presence of melt (e.g. core of grain 05G-019B-XZ-m6
Monazite exhibits strong depletions in Y + HREEs in Fig. 13d). Thus, domain 2 is regarded as a direct
when it precipitates during or after the growth of gar- record of melting and peritectic garnet growth such
net, a phase that readily incorporates Y + HREEs that Y from monazite was incorporated into garnet
(Bea & Montero, 1999; Zhu & O’Nions, 1999a). and Th from the melt was incorporated into monazite
Numerous workers have used Y-zoning in monazite to during prograde melting (stage 3 in Fig. 15). Because
infer the relative timing of garnet growth when no S1 fabrics in the Upper Deck domain are defined by
other Y-bearing phases are present in the assemblage the products of partial melting, e.g. residual garnetite
(Zhu & O’Nions, 1999b; Foster et al., 2002, 2004; Pyle layers and leucosomes (Figs 3a–d & 4a,b), the c. 2.61–
& Spear, 2003; Gibson et al., 2004; Kohn et al., 2005; 2.55 Ga dates for domain 2 broadly constrain the age
Kelly et al., 2006; Mahan et al., 2006a; Dumond of the S1 fabric and concomitant melt-weakened lower
et al., 2010). Texturally resorbed high-Y domain 1 crustal flow. This interpretation indicates a Neoarch-
monazite cores identified in both orthopyroxene-bear- ean age for Grt1 in all samples, as concluded for gran-
ing (sample 04G-097) and kyanite-bearing felsic granulites in the Northwestern (Fig. 1; Dumond et al.,
ulites (sample 05G-019B) are inferred to represent 2010; Regan et al., 2014) and Chipman domains
relicts of detrital monazite grains that existed prior to (Flowers et al., 2008; Mahan et al., 2008).
peritectic garnet growth in the felsic granulites
(Figs 12a & 13a–c). We suggest that the resorption of
Monazite as a monitor of melt extraction
domain 1 monazite marks the onset of melting and the
incorporation of Y liberated by dissolution of mon- Melt extraction has been inferred from our field
azite into peritectic garnet (stage 2 in Fig. 15). Domain observations and phase relationships. Experimental

1 Early high-Y detrital? monazite

deposited after c. 2.64-2.61 Ga
domain 1
Mnz
10 µm
2 Dissolution of high-Y monazite
Y during partial melting synchronous
with growth of pertitectic garnet
Examples: Figs 12a, 13a-c
Bt + Pl + Qtz + Mnzdomain 1 Grt + Kfs + Opx + Melt

or
Bt + Pl + Sil + Qtz + Mnzdomain 1 Grt + Kfs + Melt 10 µm domain 2
Mnz
3 Growth of Y-depleted, Th-rich

Y monazite with melt + garnet
Th
c. 2.61-2.55 Ga prograde melting
Examples: Figs 12a, 13c-d
Bt + Pl + Qtz Grt + Kfs + Opx + Melt + Mnzdomain 2

or
Bt + Pl + Sil + Qtz Grt + Kfs + Melt + Mnzdomain 2 domain 3
50 µm Mnz
4 Episodic melt loss and growth of

Th, Ca Eu (Th+Ca)-depleted, high Eu-anomaly monazite
with high-grossular garnet annuli
c. 2.55-2.54 Ga culmination of high-Grs

UHT-HP metamorphism Grt1
Examples: Figs 12a-b, 13d-e, S8
5 Melt-absent growth of Th-depleted, 20 µm

La, Ce [La+Ce]-enriched monazite with
second generation of garnet during
feldspar recrystallization low-Grs
Grt2
Fig. 15. Five-stage model for monazite c. 1.92-1.90 Ga HP-metamorphism
behaviour during melting, garnet growth and recrystallization during domain 4
and feldspar recrystallization in continental dextral transpressive strain Mnz
lower crust based on this study. Colours for Examples: Figs 12c-d, 13f-h, 14, S8
monazite domains are the same as in
Grt2 + Mnz intergrowth
Fig. S8. See text for discussion.
work has demonstrated that Th partitions preferen- (2587 39 to 2514 47 Ma, 2r; Table S4). The
tially into haplogranite melts rather than into the occurrence of domain 3 monazite in low-grossular
solid residues (Keppler & Wyllie, 1990), but also that Grt1 cores (Fig. 6a) and high-grossular Grt1 annuli
Th partitions preferentially into monazite in relation (Figs 12a,b & 13e) implies that melt extraction was
to melt (Xing et al., 2013). Consequently, melt episodic during Neoarchean prograde loading from
extraction during partial melting may leave behind a <0.8 to >1.4 GPa (Fig. 16). These results also suggest
residue that is progressively depleted in Th, and that Th-zoning in monazite from anatectic rocks may
retention of monazite during melting or fractional provide constraints on the timing of melt migration
crystallization of monazite from melt will also leave in orogenic systems (e.g. Lederer et al., 2013).
the melt depleted in Th. The sharp drop in Th that
marks the boundary between high-Th domain 2 cores
Eu-anomalies in monazite
and low-Th domain 3 rims in monazite denotes a
fundamental change in bulk composition of the sys- Positive Eu-anomalies in monazite are rare (Fig. S7;
tem prior to crystallization of domain 3 monazite see Krenn et al., 2008; Dumond et al., 2010 for
and inclusion into Grt1 (e.g. low-Th rim on grain examples). A review of the literature for which Eu
05G-019B-XZ-m6 in Figs 6a & 13d). This change is data are reported suggests monazite typically con-
most readily explained by episodic or wholesale melt tains <2300 ppm Eu and commonly exhibits negative
loss that left little Th behind in the residue for incor- Eu-anomalies on chondrite-normalized REE plots
poration into domain 3 monazite (stage 4 in Fig. 15; (Bea et al., 1994; Bea, 1996; Hermann & Rubatto,
see also Baldwin et al., 2006). Thus, melt extraction 2003; Kelly et al., 2006; Rubatto et al., 2006). Most
preceded or coincided with crystallization of domain monazite domains in this study display positive Eu-
3 monazite in samples 04G-097 and 05G-019B anomalies with Eu varying from 3092 to 4850 ppm,

Deposition of “White Gneiss” felsic granulite protolith Bt + Pl + Qtz. Textural evidence for dissolution of
c. 2.61 Ga T (°C) apatite (e.g. rounded grains included in garnet in
0 200 400 600 800 1000
0 Fig. 6a) and its absence from the felsic granulite
matrix implies that it may have contributed Ca and
0.2 And Eu to domain 2 during melting and subsequent mon-
Grt + Opx +
–10
Bt + Pl + Qtz
0.4 azite crystallization. However, domain 3 monazite
Sil
grains are Ca- and Th-depleted and inferred to have
Kfs +
Approximate Depth (km)

–20
0.6 grown during episodic melt extraction in the absence
L
Ky Grt1 cores
2.61-2.55 Ga
of apatite. Domain 3 is characterized by some of the
0.8 –30
highest positive Eu-anomalies with absolute abun-
P (GPa)
1.0 dances of Eu up to 4850 ppm (Fig. S8b; Table S5).

–40
Re-heating and Grt2
1.92-1.90 Ga
Some of these grains also occur in the high-grossular
1.2
annuli of Grt1 (Figs 12a,b & 13e) which locally pre-
1.4
–50
Grt1 annuli
serve the resorbed reactants of partial melting (Bt +
Geotherm age (Ga) Q* (mW m–2) D (km) 2.59-2.52 Ga Pl inclusions in Figs 5e & 12a,b). We attribute the
1.6 2.61-2.55 60 30 –60 large positive Eu-anomalies for domain 3 to both an
2.59-2.52 20 60
1.8
pre 1.92 20 40 increase in the modal abundance of granulite-grade
1.92-1.90 30 40
–70 garnet (sequestering Sm + Gd) and simultaneous
2.0 breakdown of plagioclase (liberating Eu for incorpo-
ration into monazite and Ca for incorporation into
Fig. 16. P–T–t path for the Upper Deck felsic granulite based
on this study. Calculated geothermal gradients are shown for garnet; stage 4 in Fig. 15). Domain 3 monazite is
comparison for a variety of crustal thicknesses (D) and mantle interpreted as a physical record of plagioclase break-
heat fluxes (Q*). The fluid-absent biotite melting field is down and HP grossular-rich garnet growth culminat-
reproduced from Vielzeuf & Montel (1994). See text for ing with UHT-HP conditions in the Neoarchean
discussion. (Figs 15 & S8c).
Monazite as a monitor of feldspar recrystallization

and several domains are below detection for Sm
(Table S5). Positive Eu-anomalies in monazite may Domain 4 monazite grains are characterized by dis-
be caused by incorporation of excess Eu into the tinct LREE enrichment, particularly La + Ce
monazite structure synchronous with breakdown of a (Fig. S8a; Table S5). Much of domain 4 is also asso-
Eu-bearing phase like plagioclase, K-feldspar or apa- ciated with recrystallized ternary feldspar or plagio-
tite. Monazite may also acquire a positive Eu-anom- clase in the matrix (Figs 12c,d, 13h & 14a), and some
aly due to depletion of available Sm + Gd in the are intergrown with or included in low-grossular gar-
system during crystallization of a modally abundant net rims (e.g. Grt2 rims of sample 05G-019B;
(Sm + Gd)-bearing phase such as garnet. Fig. 13f,g; stage 5 in Fig. 15). The weighted mean
Trace element studies reveal that granulite facies age of 1913 25 Ma (2r, MSWD = 0.1) for five
garnet generally has high Sm(N)/Gd(N), compared to domain 4 grains in sample 05G-019B constrains the
amphibolite-grade garnet (where(N) = normalized rel- timing of Grt2 growth based on inclusion and inter-
ative to chondrite; review by Jung & Hellebrand, growth of the grains in Grt2 rims (Fig. S8c). Dates
2006). Granulite-grade garnet is also generally more for elongated and asymmetric overgrowths of
enriched in Sm + Gd (13.2 ppm Sm and 18.5 ppm Domain 4 monazite in 04G-097 and 05G-026B con-
Gd for garnet granulites reported by Bea & Montero, strain the timing of dextral shearing in the Upper
1999). Modally abundant garnet could potentially Deck domain to c. 1.92–1.90 Ga (Figs 12d, 14b &
exert a strong influence on the available reservoir of S8c).
Sm + Gd + HREEs in high-grade systems. Bea & Feldspar in granulites is typically highly enriched in
Montero (1999) observed a progressive depletion of LREEs, particularly La + Ce (Bea, 1996; Villaseca
middle REEs (Sm + Gd) and HREEs in monazite et al., 2003). The marked enrichment in La + Ce
with increasing metamorphic grade and modal abun- observed in domain 4 monazite (relative to all other
dance of garnet, based on work in the Ivrea–Verbano domains plotted in Fig. S8a) is most consistent with
Kinzigite Formation. We suggest that part of the incorporation of La + Ce during recrystallization of
explanation for positive Eu-anomalies in monazite in feldspar. The observation of prominent Eu-anomalies
UHT felsic granulites of this study is due to the fact in domain 4 monazite (Fig. S8b; Table S5) is consis-
that monazite grew in the presence of modally abun- tent with dissolution–re-precipitation of pre-existing
dant garnet that sequestered Sm + Gd. high Eu domain 3 monazite during sub-solidus
Positive Eu-anomalies in lower crustal monazite growth of low-grossular Grt2 at the expense of plagio-
are further explained by evidence for breakdown of clase (stage 5 in Fig. 15). One possible interpretation
Eu-bearing phases like apatite and/or plagioclase, in is that feldspar recrystallization and growth of Grt2
this case as reactants during fluid-absent melting of concurrent with monazite dissolution–re-precipitation

was facilitated by emplacement of the c. 1.9 Ga 2010; Regan et al., 2014). This interpretation implies
Robillard granite along the western boundary of the a regional-scale (100s of km2) Neoarchean record of
Upper Deck (Fig. 2). Re-heating during dextral strain HP metamorphism (<0.8 to ≥1.4 GPa) concurrent
along the Algold Bay shear zone would have enabled with crustal thickening, intra-plating of mafic mag-
production of the syn-tectonic Grt2 coronas around mas, and lower crustal flow in all three domains of
Opx in 04G-097 (Fig. 5c), a process inferred by the EAmt (Fig. 1b).
Williams et al. (2014) to explain repeated growth of Work in the Chipman domain supports long-term
garnet in lower crustal felsic orthogneisses. Mobiliza- lower crustal residence for portions of the Athabasca
tion of alkaline fluids derived from the Robillard granulite terrane at c. 2.55–1.9 Ga (Williams & Han-
granite would also have enhanced fluid-mediated par- mer, 2006; Flowers et al., 2008; Mahan et al., 2008).
tial alteration of monazite (Hetherington & Harlov, Our results are similar to data from felsic granulites
2008; Harlov et al., 2011). Further research is in the Northwestern domain (Fig. 1b) that record
necessary to understand the relative roles of disloca- two distinct episodes of monazite growth in the
tion creep v. dissolution–precipitation creep as mech- Neoarchean and Paleoproterozoic, separated by an
anisms for recrystallizing feldspar and precipitating inferred period of isobaric cooling (Dumond et al.,
domain 4 monazite in these otherwise dry felsic gran- 2010; Regan et al., 2014). The dashed retrograde
ulites at c. 1.92–1.90 Ga (e.g. Wawrzenitz et al., path in Fig. 16 represents our interpretation for the
2012). fate of the Neoarchean Upper Deck, such that rocks
cooled to an ambient crustal geotherm after UHT-
HP crustal thickening. The c. 1.9 Ga sub-solidus
Felsic granulite P–T–t path and regional implications for
growth of a second generation of garnet in both
the Athabasca granulite terrane
04G-097 and 05G-019B is most consistent with
Figure 16 illustrates the P–T–t path inferred for re-heating and crustal reactivation during dextral
Upper Deck felsic granulites based on results from transpression at c. 1.92–1.90 Ga (Fig. 16; Dumond
this study. Prograde loading and burial of the et al., 2008). The concordant Robillard pluton (Han-
metasedimentary protolith was followed by melting mer et al., 1994; Fig. 2) and the cross-cutting granite
and growth of low-grossular peritectic garnet at 0.6– dyke dated by Baldwin et al. (2006) are evidence for
0.8 GPa and >800 °C at c. 2.61–2.55 Ga (Fig. 16). this c. 1.9 Ga re-heating event in the vicinity of the
Melting in excess of 950 °C was facilitated by Upper Deck. Flowers et al. (2006, 2008) dated the c.
emplacement of the Upper Deck mafic sill complex, 1.9 Ga record of re-heating in the adjoining Chipman
concurrent with crustal thickening and growth of domain that was facilitated by the Chipman mafic
high-grossular garnet annuli at >1.4 GPa at c. 2.59– dyke swarm (Williams et al., 1995). Dextral reactiva-
2.52 Ga (Fig. 16). The record of UHT metamor- tion and feldspar recrystallization apparently affected
phism in the Upper Deck associated with abundant large portions of the Athabasca granulite terrane at
mafic granulite sills supports its interpretation as an c. 1.9 Ga, including the EAmt and domains west of
exposed deep crustal ‘hot zone’ (e.g. Annen et al., the Grease River shear zone that were all affected by
2006). In this case, UHT metamorphism and partial the c. 1.94–1.90 Ga west-vergent Taltson arc–conti-
melting of lower crust was facilitated by under- and nent collision (Fig. 2; Dumond et al., 2008, 2010;
intra-plating of mantle-derived magma in the Bethune et al., 2013; Regan et al., 2014).
Neoarchean.
Results from this study are in stark contrast to
CONCLUSIONS
data from mafic granulite gneisses and eclogite layers
hosted within the felsic granulites of the Upper Deck Monazite in continental lower crust represents a
domain that are interpreted to represent the effects of potentially powerful monitor of partial melting, melt
HP metamorphism at c. 1.9 Ga, despite also contain- transfer, garnet growth and feldspar recrystallization.
ing zircon dated at c. 2.54–2.52 Ga (Baldwin et al., Monazite grains in this study are marked by positive
2003, 2004). We suggest that Neoarchean zircon Eu-anomalies relative to chondrite. The results sug-
dates from the mafic granulite and eclogite gneisses gest a direct link between Y, Sm, Eu and Gd in mon-
represent metamorphic growth associated with the azite and the behaviour of two major phases in
culmination of UHT-HP metamorphism identified in continental lower crust–garnet and plagioclase. Posi-
the felsic granulite gneisses in this study and consis- tive Eu-anomalies in lower crustal monazite associ-
tent with the work of Snoeyenbos et al. (1995) and ated with modally abundant garnet appear to be
Baldwin et al. (2006). This suggestion is supported by directly related to depletions of Y, Sm and Gd in
recent work on poly-metamorphic mafic and felsic monazite during/after garnet growth and enrichments
granulites in the adjoining Northwestern and Chip- of Eu (and La + Ce) in monazite during loss or
man domains that record c. 2.60–2.55 Ga HP meta- recrystallization of plagioclase. Our results from
morphism and lower crustal sub-horizontal fabric residual UHT-HP felsic granulites validate recent lab-
development that may have been coeval (Flowers oratory results that suggest monazite solubility is
et al., 2008; Mahan et al., 2008; Dumond et al., greatly reduced at high pressure (Stepanov et al.,

2012), although melt composition may still exert an laser ablation–ICP–MS study). Chemical Geology, 117, 291–
important control on this behaviour (Montel, 1993). 312.
Beaumont, C., Jamieson, R.A., Nguyen, M.H. & Lee, B.,
2001. Himalayan tectonics explained by extrusion of a low-
ACKNOWLEDGEMENTS viscosity crustal channel coupled to focused surface denuda-
tion. Nature, 414, 738–742.
C. Yakymchuk and P. Pitra are acknowledged for Beaumont, C., Jamieson, R.A., Nguyen, M.H. & Medvedev,
helpful and thorough reviews. Editor M. Brown is S., 2004. Crustal channel flows: 1. Numerical models with
applications to the tectonics of the Himalayan-Tibetan oro-
thanked for outstanding guidance during the revision gen. Journal of Geophysical Research-Solid Earth, 109, 1–29.
process. Discussions with C. L. Andronicos, J. A. doi:10.1029/2003jb002809.
Baldwin, K. H. Mahan, K. Schulmann and D. R. Bethune, K.M., Berman, R.G., Rayner, N. & Ashton, K.E.,
Snoeyenbos are greatly appreciated. C. L. Andronicos 2013. Structural, petrological and U-Pb SHRIMP
geochronological study of the western Beaverlodge domain:
first suggested we normalize the protolith bulk compo- implications for crustal architecture, multi-stage orogenesis
sition to a shale or greywacke bulk composition to test and the extent of the Taltson orogen in the SW Rae craton,
for a sedimentary precursor. Stereonets were produced Canadian Shield. Precambrian Research, 232, 89–118.
with GEOrient v. 9.5.0. Funding for this research was Brown, M., 2002. Retrograde processes in migmatites revisited.
provided by NSF-EAR grants 0310004, 0609935 and Journal of Metamorphic Geology, 20, 25–40.
Brown, M. & Solar, G.S., 1998. Shear-zone systems and melts:
1255277. P. Goncalves acknowledges the ‘defi feedback relations and self-organization in orogenic belts.
NEEDS’ from the INSU–CNRS (France) for travel Journal of Structural Geology, 20, 211–227.
support and monazite research funding. Additional Bucher, K. & Frey, M., 2002. Petrogenesis of Metamorphic
funding was provided by a 2006 Mineralogical Society Rocks. Springer, Berlin, 341 pp.
Carswell, D.A. & O’Brien, P.J., 1993. Thermobarometry and
of America Grant for Student Research in Mineralogy geotectonic significance of high-pressure granulites: examples
and Petrology to G. Dumond. Financial support for from the Moldanubian zone of the Bohemian Massif in
fieldwork was also provided by the Leo Hall and Glo- lower Austria. Journal of Petrology, 34, 427–459.
ria Radke Memorial Funds (Department of Geo- Chen, Y., Ye, K., Liu, J.-B. & Sun, M., 2008. Quantitative
sciences, University of Massachusetts) and the P–T–X constraints on orthopyroxene-bearing high-pressure
granulites in felsic-metapelitic rocks: evidence from the
University of Arkansas. Huangtuling granulite, Dabieshan Orogen. Journal of Meta-
morphic Geology, 26, 1–15.
Clemens, J.D., 2006. Melting of the continental crust: fluid
REFERENCES regimes, melting reactions, and source-rock fertility. In: Evo-
lution and Differentiation of the Continental Crust (eds
Annen, C., Blundy, J.D. & Sparks, S.J., 2006. The genesis of Brown, M. & Rushmer, T.), pp. 296–327. Cambridge
intermediate and silicic magmas in deep crustal hot zones. University Press, Cambridge.
Journal of Petrology, 47, 505–539. Clemens, J.D. & Vielzeuf, D., 1987. Constraints on melting
Baldwin, J.A., Bowring, S.A. & Williams, M.L., 2003. Petro- and magma production in the crust. Earth and Planetary
logical and geochronological constraints on high pressure, Science Letters, 86, 287–306.
high temperature metamorphism in the Snowbird tectonic Coggon, R. & Holland, T.J.B., 2002. Mixing properties of
zone, Canada. Journal of Metamorphic Geology, 21, 81–98. phengitic micas and revised garnet-phengite thermobarome-
Baldwin, J.A., Bowring, S.A., Williams, M.L. & Williams, ters. Journal of Metamorphic Geology, 20, 683–696.
I.S., 2004. Eclogites of the Snowbird tectonic zone: petro- Condie, K.C., 1993. Chemical composition and evolution of
logical and U-Pb geochronological evidence for Paleopro- the upper continental crust: contrasting results from surface
terozoic high-pressure metamorphism in the western samples and shales. Chemical Geology, 104, 1–37.
Canadian Shield. Contributions to Mineralogy and Petrology, Connolly, J.A.D., 2005. Computation of phase equilibria by
147, 528–548. linear programming: a tool for geodynamic modeling and its
Baldwin, J.A., Bowring, S.A., Williams, M.L. & Mahan, K.H., application to subduction zone decarbonation. Earth and
2006. Geochronological constraints on the evolution of high- Planetary Science Letters, 236, 524–541.
pressure felsic granulites from an integrated electron micro- Crowley, J.L., Waters, D.J., Searle, M.P. & Bowring, S.A.,
probe and ID-TIMS geochemical study. Lithos, 88, 173–200. 2009. Pleistocene melting and rapid exhumation of the
Baldwin, J.A., Powell, R., Williams, M.L. & Goncalves, P., Nanga Parbat massif, Pakistan: age and P-T conditions of
2007. Formation of eclogite, and reaction during exhumation accessory mineral growth in migmatite and leucogranite.
to mid-crustal levels, Snowbird tectonic zone, western Cana- Earth and Planetary Science Letters, 288, 408–420.
dian Shield. Journal of Metamorphic Geology, 25, 953–974. Donovan, J.J., Snyder, D.A. & Rivers, M.L., 1993. An
Baldwin, J.A., Powell, R., White, R.W. & St ıpsk
a, P., 2015. improved interference correction for trace element analysis.
Using calculated chemical potential relationships to account Microbeam Analysis, 2, 23–28.
for replacement of kyanite by symplectite in high pressure Dumond, G., McLean, N., Williams, M.L., Jercinovic, M.J. &
granulites. Journal of Metamorphic Geology, 33, 311–330. Bowring, S.A., 2008. High-resolution dating of granite petro-
Bea, F., 1996. Residence of REE, Y, Th, and U in granites genesis and deformation in a lower crustal shear zone: Atha-
and crustal protoliths: implications for the chemistry of crus- basca granulite terrane, western Canadian Shield. Chemical
tal melts. Journal of Petrology, 37, 521–552. Geology, 254, 175–196.
Bea, F. & Montero, P., 1999. Behavior of accessory phases and Dumond, G., Goncalves, P., Williams, M. L. & Jercinovic, M.
redistribution of Zr, REE, Y, Th, and U during metamor- J., 2010. Sub-horizontal fabric in exhumed continental lower
phism and partial melting of metapelites in the lower crust: an crust and implications for lower crustal flow: Athabasca
example from the Kinzigite Formation of Ivrea-Verbano, NW granulite terrane, western Canadian Shield. Tectonics, 29,
Italy. Geochimica et Cosmochimica Acta, 63, 1133–1153. TC2006. doi:10.1029/2009TC002514.
Bea, F., Pereira, M.D. & Stroh, A., 1994. Mineral-leucosome Dumond, G., Mahan, K.H., Williams, M.L. & Jercinovic, M.J.,
trace element partitioning in a peraluminous migmatite (a 2013. Transpressive uplift and exhumation of continental

lower crust revealed by synkinematic monazite reactions. dissolution-reprecipitation in element redistribution and mass
Lithosphere, 5, 507–512. transfer. Contributions to Mineralogy and Petrology, 162,
Fiala, J., Matejovska, O. & Va nkova, V., 1987. Moldanubian 329–348.
granulites: source material and petrogenetic considerations. Hermann, J. & Rubatto, D., 2003. Relating zircon and mon-
Neues Jahrbuch f€ ur Mineralogie Abhandlungen, 157, 133–165. azite domains to garnet growth zones: age and duration of
Flowers, R., Bowring, S. & Williams, M., 2006. Timescales granulite facies metamorphism in the Val Malenco lower
and significance of high-pressure, high-temperature meta- crust. Journal of Metamorphic Geology, 21, 833–852.
morphism and mafic dike anatexis, Snowbird tectonic zone, Hetherington, C.J. & Harlov, D.E., 2008. Metasomatic thorite
Canada. Contributions to Mineralogy and Petrology, 151, and uraninite inclusions in xenotime and monazite from
558–581. granitic pegmatites, Hidra anorthosite massif, southwestern
Flowers, R.M., Bowring, S., Mahan, K.H. & Williams, M.L., Norway: mechanics and fluid chemistry. American Mineralo-
2008. Stabilization and reactivation of cratonic lithosphere gist, 93, 806–820.
from the lower crustal record in the western Canadian Holland, T.J.B. & Powell, R., 1998. An internally consistent
shield. Contributions to Mineralogy and Petrology, 156, 529– thermodynamic data set for phases of petrological interest.
549. Journal of Metamorphic Geology, 16, 309–343.
Foster, G., Gibson, H.D., Parrish, R., Horstwood, M., Fraser, Hollister, L.S. & Crawford, M.L., 1986. Melt-enhanced defor-
J. & Tindle, A., 2002. Textural, chemical and isotopic mation: a major tectonic process. Geology, 14, 558–561.
insights into the nature and behaviour of metamorphic mon- Jercinovic, M.J. & Williams, M.L., 2005. Analytical perils (and
azite. Chemical Geology, 191, 183–207. progress) in electron microprobe trace element analysis
Foster, G., Parrish, R.R., Horstwood, M.S.A., Chenery, S., applied to geochronology: background acquisition, interfer-
Pyle, J. & Gibson, H.D., 2004. The generation of prograde ences, and beam irradiation effects. American Mineralogist,
P-T–t points and paths; a textural, compositional, and 90, 526–546.
chronological study of metamorphic monazite. Earth and Jercinovic, M.J., Williams, M.L. & Lane, E.D., 2008. In-situ
Planetary Science Letters, 228, 125–142. trace element analysis of monazite and other fine-grained
Frost, B.R., Barnes, C.G., Collins, W.J., Arculus, R.J., Ellis, accessory minerals by EPMA. Chemical Geology, 254, 197–
D.J. & Frost, C.D., 2001. A geochemical classification for 215.
granitic rocks. Journal of Petrology, 42, 2033–2048. Johnson, T.E., White, R.W. & Powell, R., 2008. Partial melt-
Fuhrman, M.L. & Lindsley, D.H., 1988. Ternary-feldspar ing of metagreywacke: a calculated mineral equilibria study.
modeling and thermometry. American Mineralogist, 73, 201– Journal of Metamorphic Geology, 26, 837–853.
215. Jung, S. & Hellebrand, E., 2006. Trace element fractionation
Gasquet, D., Bertrand, J.-M., Paquette, J.-L. et al., 2010. Mio- during high-grade metamorphism and crustal melting—con-
cene to Messinian deformation and hydrothermal activity in straints from ion microprobe data of metapelitic, migmatitic
a pre-Alpine basement massif of the French western Alps: and igneous garnets and implications for Sm–Nd garnet
new U–Th–Pb and Ar ages from the Lauziere massif. Bul- chronology. Lithos, 87, 193–213.
letin de la Soci
et
e Geologique de France, 181, 227–241. Kelly, N.M., Clarke, G.L. & Harley, S.L., 2006. Monazite
Gibson, H.D., Carr, S.D., Brown, R.L. & Hamilton, M.A., behaviour and age significance in poly-metamorphic high-
2004. Correlations between chemical and age domains in grade terrains: a case study from the western Musgrave
monazite, and metamorphic reactions involving major pelitic Block, central Australia. Lithos, 88, 100–134.
phases: an integration of ID–TIMS and SHRIMP Kelsey, D.E. & Hand, M., 2015. On ultrahigh temperature crus-
geochronology with Y-Th–U X–ray mapping. Chemical tal metamorphism: Phase equilibria, trace element thermome-
Geology, 211, 237–260. try, bulk composition, heat sources, timescales, and tectonic
Goncalves, P., Nicollet, C. & Montel, J.-M., 2004. Petrology settings. Geoscience Frontiers, 6, 311–356. doi:10.1016/j.gsf.
and in situ U–Th–Pb monazite geochronology of ultrahigh- 2014.09.006.
temperature metamorphism from the Andriamena mafic Kelsey, D.E., Clark, C. & Hand, M., 2008. Thermobarometric
unit, North-Central Madagascar. Significance of a petro- modelling of zircon and monazite growth in melt-bearing
graphical P–T path in a polymetamorphic context. Journal systems: examples using model metapelitic and metapsam-
of Petrology, 45, 1923–1957. mitic granulites. Journal of Metamorphic Geology, 26, 199–
Green, D.H. & Lambert, I.B., 1965. Experimental crystalliza- 212.
tion of anhydrous granite at high pressures and tempera- Keppler, H. & Wyllie, P.J., 1990. Role of fluids in transport
tures. Journal of Geophysical Research, 70, 5259–5268. and fractionation of uranium and thorium in magmatic pro-
Green, E.C.R., Holland, T.J.B. & Powell, R., 2007. An order– cesses. Nature, 348, 531–533.
disorder model for omphacitic pyroxenes in the system Kohn, M.J., Wieland, M.S., Parkinson, C.D. & Upreti, B.N.,
jadeite–diopside–hedenbergite–acmite, with applications to 2005. Five generations of monazite in Langtang gneisses:
eclogite rocks. American Mineralogist, 92, 1181–1189. implications for chronology of the Himalayan metamorphic
Hanmer, S., 1994. Geology, East Athabasca Mylonite Triangle, core. Journal of Metamorphic Geology, 23, 399–406.
Saskatchewan: Map 1859A, Scale 1:100000. Geological Sur- Kopf, C.F., 1999. Deformation, Metamorphism, and Magma-
vey of Canada, Ottawa. tism in the East Athabasca Mylonite Triangle, Northern Sas-
Hanmer, S., 1997. Geology of the Striding-Athabasca mylonite katchewan: Implications for the Archean and Early
zone, northern Saskatchewan and southeastern District of Proterozoic Crustal Structure of the Canadian Shield. Unpub-
Mackenzie, Northwest Territories. Geological Survey of lished thesis, University of Massachusetts, Amherst, MA,
Canada Bulletin, 501, 92 pp. 156 pp.
Hanmer, S., Parrish, R.R., Williams, M.L. & Kopf, C., 1994. Korhonen, F.J., Brown, M., Clark, C. & Bhattacharya, S.
Striding-Athabasca mylonite zone: complex Archean deep 2013. Osmulite–melt interactions in ultrahigh temperature
crustal deformation in the East Athabasca mylonite triangle, granulites: phase equilibria modelling and implications for
N. Saskatchewan. Canadian Journal of Earth Sciences, 31, the P–T–t evolution of the Eastern Ghats Province, India.
1287–1300. Journal of Metamorphic Geology, 31, 881–907.
Hanmer, S., Williams, M.L. & Kopf, C., 1995. Modest move- Korhonen, F.J., Clark, C., Brown, M. & Taylor, R.J.M., 2014.
ments, spectacular fabrics in an intracontinental deep-crustal Taking the temperature of earth’s hottest crust. Earth and
strike-slip fault: striding-Athabasca mylonite zone, NW Planetary Science Letters, 408, 341–354.
Canadian Shield. Journal of Structural Geology, 17, 493–507. Krenn, E., Putz, H., Finger, F. & Paar, W., 2008. Unusual
Harlov, D.E., Wirth, R. & Hetherington, C.J., 2011. Fluid- monazite with high S, Sr, Eu, and common Pb contents in
mediated partial alteration in monazite: the role of coupled ore bearing mylonites from the Schellgaden mining district,

Austria. Geophysical Research Abstracts, 10, EGU2008-A- Pyle, J.M., Spear, F.S. & Wark, D.A., 2002. Electron micro-
11796. probe analysis of REE in Apatite, monazite, and xenotime:
Krikorian, L., 2002. Geology of the Wholdaia Lake segment of protocols and pitfalls. In: Reviews in Mineralogy and Geo-
the Snowbird Tectonic Zone, Northwest Territories (Nu- chemistry (eds Kohn, M.J., Rakovan, J. & Hughes, J.M.),
navut): A View of the Deep Crust During Assembly and Sta- pp. 337–362. Mineralogical Society of America, Washington,
bilization of the Laurentian craton. Unpublished MS thesis, DC.
University of Massachusetts, Amherst, MA, 90 pp. Pyle, J.M., Spear, F.S., Wark, D.A., Daniel, C.G. & Storm,
Lederer, G.W., Cottle, J.M., Jessup, M.J., Langille, J.M. & L.C., 2005. Contributions to precision and accuracy of
Ahmad, T., 2013. Timescales of partial melting in the Hima- chemical ages of monazite. American Mineralogist, 90, 547–
layan middle crust: insight from the Leo Pargil dome, north- 577.
west India. Contributions to Mineralogy and Petrology, 166, Rapp, R.P. & Watson, E.B., 1986. Monazite solubility and dis-
1415–1441. solution kinetics: implications for the Th and light rare-earth
Mahan, K.H. & Williams, M.L., 2005. Reconstruction of a chemistry of felsic magmas. Contributions to Mineralogy and
large deep-crustal terrane: implications for the Snowbird tec- Petrology, 94, 304–316.
tonic zone and early growth of Laurentia. Geology, 33, 385– Rapp, R.P., Ryerson, F.J. & Miller, C.F., 1987. Experimental
388. evidence bearing on the stability of monazite during crustal
Mahan, K.H., Williams, M.L. & Baldwin, J.A., 2003. Contrac- anatexis. Geophysical Research Letters, 14, 307–310.
tional uplift of deep crustal rocks along the Legs Lake shear Regan, S., Williams, M.L., Mahan, K.H., Leslie, S., Jercinovic,
zone, western Churchill Province, Canadian Shield. Canadian M.J. & Holland, M.E., 2014. The Cora Lake shear zone,
Journal of Earth Sciences, 40, 1085–1110. Athabasca granulite terrane: an intraplate response to far-
Mahan, K.H., Goncalves, P., William, M.L. & Jercinovic, field orogenic processes during the amalgamation of Lauren-
M.J., 2006a. Dating metamorphic reactions and fluid flow: tia. Canadian Journal of Earth Sciences, 51, 877–901.
application to exhumation of high-P granulites in a crustal- Reno, B.L., Piccoli, P.M., Brown, M. & Trouw, R.A.J., 2012.
scale shear zone, western Canadian Shield. Journal of Meta- In situ monazite (U–Th)–Pb ages from the Southern Brasilia
morphic Geology, 24, 193–217. Belt, Brazil: constraints on the high-temperature retrograde
Mahan, K.H., Williams, M.L., Flowers, R.M., Jercinovic, evolution of HP granulites. Journal of Metamorphic Geology,
M.J., Baldwin, J.A. & Bowring, S.A., 2006b. Geochronologi- 30, 81–112.
cal constraints on the Legs Lake shear zone with implica- Rosenberg, C.L. & Handy, M.R., 2005. Experimental deforma-
tions for regional exhumation of lower continental crust, tion of partially melted granite revisited: implications for the
western Churchill Province, Canadian Shield. Contributions continental crust. Journal of Metamorphic Geology, 23, 19–28.
to Mineralogy and Petrology, 152, 223–242. Rubatto, D., Hermann, J. & Buick, I.S., 2006. Temperature
Mahan, K.H., Goncalves, P., Flowers, R., Williams, M.L. & and bulk composition control on the growth of monazite
Hoffman-Setka, D., 2008. The role of heterogeneous strain and zircon during low-pressure anatexis (Mount Stafford,
in the development and preservation of a polymetamorphic central Australia). Journal of Petrology, 47, 1973–1996.
record in high-P granulites, western Canadian Shield. Jour- Rudnick, R.L., 1992. Restites, Eu-anomalies, and the lower con-
nal of Metamorphic Geology, 26, 669–694. tinental crust. Geochimica et Cosmochimica Acta, 56, 963–970.
Mahan, K.H., Smit, C.A., Williams, M.L., Dumond, G. & van Rudnick, R.L. & Gao, S., 2003. Composition of the continen-
Reenen, D.D., 2011. Heterogeneous strain and polymeta- tal crust. In: Treatise on Geochemistry (eds Holland, H.D. &
morphism in high-grade terranes: insight into crustal pro- Turekian, K.K.), Vol. 3, pp. 1–64. Elsevier, Oxford.
cesses from the Athabasca Granulite Terrane, western Searle, M., 2013. Crustal melting, ductile flow, and deforma-
Canada, and the Limpopo Complex, southern Africa. In: tion in mountain belts: cause and effect relationships. Litho-
Origin and Evolution of High-grade Gneiss Terranes, with sphere, 5, 547–554.
Special Emphasis on the Limpopo Complex of Southern Africa Skora, S. & Blundy, J., 2012. Monazite solubility in hydrous
(eds van Reenen, D.D., Kramers, J.D., McCourt, S. & Per- silicic melts at high pressure conditions relevant to subduc-
chuk, L.L.), Geological Society of America Memoir, 207, tion zone metamorphism. Earth and Planetary Science Let-
269–287. ters, 321–322, 104–114.
Martel, E., van Breemen, O., Berman, R.G. & Pehrsson, S., Slimmon, W.L., 1989. Bedrock compilation geology - Fond du
2008. Geochronology and tectonometamorphic history of the Lac (NTS 74–O): Map 247A, scale 1:250000, Saskatchewan
Snowbird Lake area, Northwest Territories, Canada: new Geological Survey, Saskatchewan Energy and Mines.
insights into the architecture and significance of the Snowbird Snoeyenbos, D.R., Williams, M.L. & Hanmer, S., 1995.
tectonic zone. Precambrian Research, 161, 201–230. Archean high-pressure metamorphism in the western Cana-
Martins, L., Vlach, S.R.F. & Janasi, V., 2009. Reaction micro- dian Shield. European Journal of Mineralogy, 7, 1251–1272.
textures of monazite: correlation between chemical and age Spear, F.S. & Kohn, M.J., 1996. Trace element zoning in gar-
domains in the Nazare Paulista migmatite, SE Brazil. Chemi- net as a monitor of crustal melting. Geology, 24, 1099–1102.
cal Geology, 261, 271–285. Stepanov, A.S., Hermann, J., Rubatto, D. & Rapp, R.P.,
Montel, J.M., 1986. Experimental determination of the solubil- 2012. Experimental study of monazite/melt partitioning with
ity of Ce-monazite in SiO2–Al2O3–K2O–Na2O melts at implications for the REE, Th and U geochemistry of crustal
800°C, 2 kbar, under H2O-saturated conditions. Geology, 14, rocks. Chemical Geology, 300–301, 200–220.
659–662. ıpsk
St a, P. & Powell, R., 2005. Does ternary feldspar constrain
Montel, J.-M., 1993. A model for monazite/melt equilibrium metamorphic conditions of high-grade meta-igneous rocks?
and application to the generation of granitic magmas. Chem- Evidence from orthopyroxene granulites, Bohemian Massif.
ical Geology, 110, 127–146. Journal of Metamorphic Geology, 23, 627–647.
O’Brien, P.J. & R€ otzler, J., 2003. High-pressure granulites: for- Sun, S.-S. & McDonough, W.F., 1989. Chemical and isotope
mation, recovery of peak conditions and implications for tec- systematics of oceanic basalts: implications for mantle com-
tonics. Journal of Metamorphic Geology, 21, 3–20. positions and processes. In: Magmatism in the Ocean Basins
Pyle, J.M. & Spear, F.S., 2003. Four generations of accessory- (eds Saunders, A.D. & Norry, M.J.), pp. 313–345. Geologi-
phase growth in low-pressure migmatites from SW New cal Society, London.
Hampshire. American Mineralogist, 88, 338–351. Tajcmanov a, L., Connolly, J.A.D. & Cesare, B., 2009. A ther-
Pyle, J.M., Spear, F.S., Rudnick, R.L. & McDonough, W.F., modynamic model for titanium and ferric iron solution in
2001. Monazite-xenotime-garnet equilibrium in metapelites biotite. Journal of Metamorphic Geology, 27, 153–165.
and a new monazite-garnet thermometer. Journal of Petrol- Tropper, P., Knozett, J. & Finger, F., 2005. Experimental con-
ogy, 42, 2083–2107. straints on the formation of high-P/high-T granulites in the

Southern Bohemian Massif. European Journal of Mineralogy, reporting microprobe monazite ages. Chemical Geology, 225,
17, 343–356. 1–15.
Vielzeuf, D. & Montel, J.M., 1994. Partial melting of meta- Williams, M.L., Jercinovic, M.J. & Hetherington, C.J., 2007.
greywackes. Part I. Fluid-absent experiments and phase rela- Microprobe monazite geochronology: understanding geo-
tionships. Contributions to Mineralogy and Petrology, 117, logic processes by integrating composition and chronology.
375–393. Annual Review of Earth and Planetary Sciences, 35, 137–175.
Vielzeuf, D. & Schmidt, M.W., 2001. Melting relations in Williams, M.L., Karlstrom, K.E., Dumond, G. & Mahan,
hydrous systems revisited: application to metapelites, meta- K.H., 2009. Perspectives on the architecture of continental
greywackes and metabasalts. Contributions to Mineralogy crust from integrated field studies of exposed isobaric sec-
and Petrology, 141, 251–267. tions. Geological Society of America Special Papers, 456,
Villaseca, C., Romera, C.M., De la Rosa, J. & Barbero, L., 219–241.
2003. Residence and redistribution of REE, Y, Zr, Th and Williams, M.L., Dumond, G., Mahan, K.H., Regan, S.P. &
U during granulite-facies metamorphism: behaviour of acces- Holland, M., 2014. Garnet-forming reactions in felsic
sory and major phases in peraluminous granulites of central orthogneiss: implications for densification and strengthening
Spain. Chemical Geology, 200, 293–323. of the lower continental crust. Earth and Planetary Science
Watt, G.R., 1995. High-thorium monazite-(Ce) formed during Letters, 405, 207–219.
disequilibrium melting of metapelites under granulite-facies Xing, L., Trail, D. & Watson, B.E., 2013. Th and U partition-
conditions. Mineralogical Magazine, 59, 735–743. ing between monazite and felsic melt. Chemical Geology,
Watt, G.R. & Harley, S.L., 1993. Accessory phase controls on 358, 46–53.
the geochemistry of crustal melts and restites produced dur- Yakymchuk, C. & Brown, M., 2014a. Consequences of open-
ing water-undersaturated partial melting. Contributions to system melting in tectonics. Journal of the Geological Society,
Mineralogy and Petrology, 114, 550–566. 171, 21–40.
Wawrzenitz, N., Krohe, A., Rhede, D. & Romer, R.L., 2012. Yakymchuk, C. & Brown, M., 2014b. Behaviour of zircon and
Dating rock deformation with monazite: the impact of disso- monazite during crustal melting. Journal of the Geological
lution precipitation creep. Lithos, 134–135, 52–74. Society, 171, 465–479.
€
Weiss, C.S., 1803. Uber die Gebirgsart des s€achsischen Erzge- Yakymchuk, C., Brown, M., Clark, C. et al., 2015. Decoding
birges, welche unter dem Namen Weiss-Stein neuerlich polyphase migmatites using geochronology and phase equilib-
bekannt gemacht worden ist. Neue Schriften der Gesellschaft ria modeling. Journal of Metamorphic Geology, 33, 203–230.
naturforschender Freunde zu Berlin, 4, 342–366. Zhu, X.K. & O’Nions, R.K., 1999a. Monazite chemical com-
White, R.W. & Powell, R., 2002. Melt loss and the preserva- position: some implications for monazite geochronology.
tion of granulite facies mineral assemblages. Journal of Contributions to Mineralogy and Petrology, 137, 351–363.
Metamorphic Geology, 20, 621–632. Zhu, X.K. & O’Nions, R.K., 1999b. Zonation of monazite in
White, R.W., Powell, R., Holland, T.J.B. & Worley, B.A., metamorphic rocks and its implications for high temperature
2000. The effect of TiO2 and Fe2O3 on metapelitic assem- thermochronology: a case study from the Lewisian terrain.
blages at greenschist and amphibolite facies conditions: min- Earth and Planetary Science Letters, 171, 209–220.
eral equilibria calculations in the system K2O–FeO–MgO–
Al2O3–SiO2–H2O–TiO2–Fe2O3. Journal of Metamorphic
Geology, 18, 497–511. SUPPORTING INFORMATION
White, R.W., Powell, R. & Holland, T.J.B., 2001. Calculation
of partial melting equilibria in the system Na2O–CaO–K2O– Additional Supporting Information may be found in
FeO–MgO–Al2O3–SiO2–H2O (NCKFMASH). Journal of the online version of this article at the publisher’s
Metamorphic Geology, 19, 139–153. web site:
White, R.W., Powell, R. & Halpin, J.A., 2004. Spatially-focused
melt formation in aluminous metapelites from Broken Hill,
Figure S1. Backscattered electron image of garnet
Australia. Journal of Metamorphic Geology, 20, 621–632. in Opx-felsic granulite 04G-097. Note the abundant
White, R.W., Powell, R. & Holland, T.J.B., 2007. Progress Fe-sulphide inclusions (pyrrhotite-Po) and associated
relating to calculation of partial melting equilibria for meta- inclusions of Bt + Pl + Rt. Ilmenite was notably
pelites. Journal of Metamorphic Geology, 25, 511–527. absent in garnet porphyroblasts and only observed in
Williams, M.L. & Hanmer, S., 2006. Structural and metamor-
phic processes in the lower crust: evidence from a deep-crus- the matrix.
tal isobarically cooled terrane, Canada. In: Evolution and Figure S2. Plane-polarized light photomicrograph
Differentiation of the Continental Crust (eds Brown, M. & of garnet porphyroblast in (kyanite + sillimanite)-
Rushmer, T.), pp. 231–267. Cambridge University Press, bearing felsic granulite 05G-019B illustrating rounded
Cambridge.
Williams, M.L. & Jercinovic, M.J., 2002. Microprobe monazite
inclusions of quartz with graphite. Graphite also
geochronology: putting absolute time into microstructural occurs as an abundant fabric-defining phase in the
analysis. Journal of Structural Geology, 24, 1013–1028. matrix.
Williams, M.L. & Jercinovic, M.J., 2012. Tectonic interpreta- Figure S3. High-resolution full thin-section Ca Ka
tion of metamorphic tectonites: integrating compositional map of D2 ultramylonite sample 05G-026B illustrat-
mapping, microstructural analysis and in situ monazite dat-
ing. Journal of Metamorphic Geology, 30, 739–752. ing setting of monazite grains (Fig. 14). Note the
Williams, M.L., Hanmer, S., Kopf, C. & Darrach, M., 1995. dextral sense of shear defined by fractured garnet
Syntectonic generation and segregation of tonalitic melts porphyroclasts mantled by tails of quartz ribbons.
from amphibolite dikes in the lower crust, Striding-Atha- Figure S4. Whole-rock geochemistry for Upper
basca mylonite zone, northern Saskatchewan. Journal of
Geophysical Research, 100(B8), 15717–15734.
Deck felsic granulites and protolith from this study
Williams, M.L., Melis, E.A., Kopf, C. & Hanmer, S., 2000. Mi- plotted with data from Baldwin et al. (2006). Compo-
crostructural tectonometamorphic processes and the develop- sitions for greywacke (Condie, 1993; Vielzeuf &
ment of gneissic layering: a mechanism for metamorphic Montel, 1994), shale (Condie, 1993), high-pressure
segregation. Journal of Metamorphic Geology, 18, 41–57. granite (Green & Lambert, 1965) and the Bohemian
Williams, M.L., Jercinovic, M.J., Goncalves, P. & Mahan, K.,
2006. Format and philosophy for collecting, compiling, and Massif felsic granulites (Fiala et al., 1987; Carswell &

O’Brien, 1993; St ıpska & Powell, 2005; Tropper Figure S8. (a) XLaPO4 þ CePO4 v: XCheralite for all
et al., 2005) are plotted for comparison. (a) Fe# v. analysed monazite domains in this study. Mole frac-
SiO2 (Frost et al., 2001). (b) Modified alkali lime tions for components in monazite were calculated
index v. SiO2 (Frost et al., 2001). (c) Protolith bulk according to Pyle et al. (2001). Arrows connect core
composition (04G-090A) normalized against average and rim analyses from the same grain, e.g. domain 2
Late Archean greywacke and average Archean cra- to domain 3. Note core-to-rim enrichment in La +
tonic shale of Condie (1993). (d) CaO v. K2O. (e) Ce. (b) XEuPO4 v: XCheralite for all analysed monazite
AFM diagram projected from Kfs + Qtz. domains in this study. Note core-to-rim enrichment
Figure S5. Rare earth element geochemistry for fel- in Eu. (c) Summary of all EPMA monazite dates
sic granulite protolith, dyke and felsic granulites from (2r) for samples 04G-097, 05G-026B and 05G-
this study and Baldwin et al. (2006). All data are 019B, and the Moacir consistency standard. All his-
normalized to the composition of bulk continental tograms represent normal Gaussian distributions
crust from Rudnick & Gao (2003). scaled relative to the consistency standard (after Wil-
Figure S6. Temperature v. XFe2 O3 pseudosection liams et al., 2006). The rationale behind weighted
calculated at 1.45 GPa with Perple_X 6.6.8. Note means for domains 3 and 4 is presented in the discus-
‘Rutile out’ at the far left of both diagrams. (a) sion.
T XFe2 O3 using the melt reintegrated ‘protolith’ bulk Table S1. Whole-rock major and rare earth ele-
composition for the Opx-bearing felsic granulite in ment geochemistry for Upper Deck felsic granulites
Fig. 8 and Table S4 (sample 04G-097). (b) T XFe2 O3 and protolith.
using the melt reintegrated ‘protolith’ bulk composi- Table S2. Mineral compositions for orthopyrox-
tion for the (Sil + Ky)-bearing felsic granulite in ene-bearing and aluminosilicate (kyanite + silliman-
Fig. 11 and Table S4 (sample 04G-097). Reducing ite)-bearing felsic granulites.
conditions (XFe2 O3 0.05) are required to produce Table S3. Solution models used in Perple_X 6.6.8.
the peak rutile-bearing assemblages observed in this Table S4. Melt compositions and melt reintegrated
study. bulk compositions derived via Perple_X 6.6.8 for the
Figure S7. Chondrite-normalized rare earth ele- 4-stage melt reintegration models illustrated in Figs 7
ment diagram for monazite grains analysed in this and 11.
study (denoted by solid black lines). Values for Tb– Table S5. Electron microprobe major and trace ele-
Lu in monazite data via EPMA from this study were ment data for monazite.
below detection. Monazite data from Bea (1996) are
shown for comparison. Chondrite values are from Received 27 August 2014; revision accepted 28 June 2015.
Sun & McDonough (1989).

Dumond 2015

Uploaded by

Copyright:

Available Formats

Dumond 2015

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dumond 2015

Uploaded by

Copyright:

Available Formats

J. metamorphic Geol., 2015, 33, 735–762 doi:10.1111/jmg.

Monazite as a monitor of melting, garnet growth and feldspar

The ability to accurately date the processes involved

© 2015 John Wiley & Sons Ltd 735

(a) 60 ° (b) Whold

1.0-1.2 GPa 8 Felsic granulite 1.3 GPa

775-860 ºC >1.5 0.9-1.0 GPa 850 ºC

© 2015 John Wiley & Sons Ltd

eclogite, c. 1.9 Ga Robillard 72

garnet clinopyroxenite granite (Hanmer 12 streching/intersection lineation upright F2 synform

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

synchronous with the S1 sub-horizontal fabric.

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

(a) 400 µm (b) 200 µm (c)

Pl Sil-bearing felsic granulite 05G-019B illustrating setting of

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

04G-097: Opx-bearing felsic granulite

that a melt connectivity threshold of 1.2 Opx Bt 2 Opx L Kfs

melt reintegration stages (numbers 1–4) Pl Grt

pseudosection in Fig. 8. The composition of 0.8 4

© 2015 John Wiley & Sons Ltd

04G-097: Opx-bearing felsic granulite

stage melt reintegration procedure (Fig. 7;

Table S4). The black arrow corresponds to

Opx L Kfs Ilm

0.6 reintegration was performed. The numbers

Fig. 9 is denoted by the star. (b) High-T

side of same P–T pseudosection in (a) with

© 2015 John Wiley & Sons Ltd

04G-097: Opx-bearing felsic granulite

Opx L Kfs Grt Rt

Opx L Kf Pl Grt Rt 0.12

Opx Opx out

900 Kfs Bt Bt L Kfs Grt

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

05G-019B: Sil + Ky-bearing felsic granulite

to achieve ~1 mol.% melt at the inferred L

7 mol.% melt (~7 vol.%) was reached at

(numbers 1–4) were performed at the L Kfs Pl 3

of melt and the reintegrated bulk 0.8 4

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

© 2015 John Wiley & Sons Ltd

(a) 04G-097-2-m1 [inclusion in Grt1] Grt1

Grt1 Grt1 04G-097-2-m7

© 2015 John Wiley & Sons Ltd

(a) (b) 05G-019B-YZ-2-m12 (c) 05G-019B-XZ-2-m11

(d) (e) Domain 3 rim

© 2015 John Wiley & Sons Ltd

2 monazite cores and overgrowths are depleted in Y

© 2015 John Wiley & Sons Ltd

1 Early high-Y detrital? monazite

Bt + Pl + Qtz + Mnzdomain 1 Grt + Kfs + Opx + Melt