1
Contribution of AMS studies in understanding the migmatitic terrains
2
of Pointe Géologie, Terre Adélie (East-Antarctica)
3
Jérôme Bascou1, Bernard Henry2, René-Pierre Ménot1, Minoru Funaki3, Guilhem Barruol4
4
5
6
7
8
9
10
11
1
Université de Lyon, Université Jean Monnet and UMR-CNRS 6524, Laboratoire Magmas et Volcans, 23
rue du Dr Paul Michelon, 42023 Saint Etienne, France
2
Paléomagnétisme, Institut de Physique du Globe de Paris, Sorbonne Paris Cité, Univ ersité Paris Diderot
and UMR-CNRS 7154, France
3
National Institute of Polar Research, 10-3 Midori-cho Tachikawa Tokyo 190-8518, Japan
4
Université de la Réunion, CNRS and Institut de Physique du Globe de Paris, Géosciences Réunion, 97715
Saint Denis cedex 9, La Réunion, France
12
13
14
15
16
17
18
19
20
Abstract
21
22
Keywords: Migmatite, anatexite, AMS, Pointe Géologie, Antarctique
23
1
24
1. Introduction
25
26
Anatexites result from progressive melting of rocks under high-grade metamorphism
27
conditions. These rocks are often the only key to determine the structural conditions in the
28
deep crust. However, the structural analysis of migmatites is often very difficult because of
29
their complex and multiphase evolution. Moreover, the products from the melting of an initial
30
highly deformed rock have a fabric that mostly cannot be determined by the classical
31
structural approaches. In partially melted gneisses, this fabric can differ from the initial one,
32
and in dominantly melted rocks the single visible fabric is often related to the melanosome
33
orientation and therefore offers an incomplete signature of tectonic events suffered by the
34
rocks. One of the characteristics of the migmatite structures is their local variability and their
35
analysis then passes by a statistical approach. To this aim, the field observations are limited to
36
part of the elements of the fabric and generally not well-adapted.
37
The anisotropy of magnetic susceptibility (AMS) technique (Graham, 1954); see
38
(Tarling and Hrouda, 1993; Borradaile and Henry, 1997; Borradaile and Jackson, 2004) and
39
references therein) has been widely used for structural applications. It offers several
40
advantages relatively to field observations: It gives not only foliation and lineation but also
41
intensity and shape of the fabric and can be applied to a set of samples representative of a
42
significant area allowing simple statistical analysis. Ferré et al. (2003) and Kruckenberg et al.
43
(2010) applied this magnetic tool for migmatites studies. They showed that despite
44
rheological variations during deformation (solid-state deformation followed by viscous flow),
45
AMS axes were consistent with observed regional field structure and thus highlighting the
46
interest of the AMS data for the strain characterizations in high-temperature metamorphic
47
rocks. In paramagnetic migmatites, the biotite fabric can evidenced these two types of
48
deformation (Hasalovà et al., 2008). Others authors have shown that anatexites could have
49
specific AMS signature, associated with changes towards weak solid-state rheology and that
50
AMS characterizations could therefore help to better model the different stages of migmatitic
51
terranes (Charles et al., 2009; Schulmann et al., 2009; Kratinová et al., 2012).
52
The Pointe Géologie archipelago (Terre Adélie, East Antarctica) represents a deep and
53
hot crust that experimented anatectic event during Paleoproterozoic times, 1.69 Ga ago
54
(Ménot et al., 2007 and references therein). This archipelago is partially covered by the ice
55
cap and the limited outcrops are key zones to constrain the Terre Adélie Craton tectonics.
56
Pelletier et al. (2002) and Gapais et al. (2008) proposed a detailed structural and kinematic
57
analyses of the area based on foliation and lineation measurements of migmatites. But we
2
58
believe these structural data may be supplemented by taking into account the magnetic fabric
59
of the leucosomes and anatexites, which correspond to more than 50% of the outcrops and
60
which are devoid of clear markers of deformation. This would greatly improve the knowledge
61
of the structural pattern of this region.
62
The first aims of this paper is thus to specify the deformation framework associated to
63
the Paleoproterozoic event in the Terre Adélie Craton from an enlarge sampling in the Pointe
64
Géologie rocks. For this, a detailed AMS study was performed in the migmatites, the larger
65
granitic dykes and the anatexites observed on the field. Finally, we discuss on the
66
relationships between the various observed migmatitic structures of centimetric to decametric
67
scale, which correspond to various degree of segregation and transportation of magma, the
68
AMS signature and the rheological evolution of the molten crust.
69
70
2. Geological setting
71
72
The Terre Adélie Craton (TAC) is located on the easternmost part of the East
73
Antarctic Shield. It crops out as islands and capes scattered along the coast between 135° and
74
145°E but most of the outcrops are located in three main archipelagos, from West to East:
75
around Pointe Géologie (66°50’S, 140°E), East of Port Martin (66°80’S, 142°25’E) and East
76
of Cape Denison (67°S, 142°25’E) up to the Mertz glacier. This crystalline basement consists
77
of Late Archean and Paleoproterozoic formations devoid of any significant reworking since
78
1.5 Ga and then it is considered as part of the Mawson block or Mawson continent (Fanning
79
et al., 1995; 2002). At the scale of the TAC, the Late Archean and Paleoproterozoic terrains
80
are distributed westwards and eastwards from Port Martin, respectively (Monnier et al., 1996;
81
Ménot et al., 2007), Figure 1.
82
This paper deals only with the Paleoproterozoic formations of the Pointe Géologie area,
83
i.e. close to the permanent French Polar Station Dumont d’Urville (DDU). As described by
84
(Bellair, 1961b, a), the main rock types are migmatitic to anatectic gneisses that are cross cut
85
by dykes of apparently isotropic coarse-grained pink granites. The detailed mapping, carried
86
on during the GEOLETA project, improves this previous description and points out some new
87
and significant features (Monnier et al., 1996; Peucat et al., 1999; Gapais et al., 2008).
88
Migmatitic gneisses and anatexites derived from a sedimentary sequence dominated by
89
aluminous pelites together with subordinate greywackes, quartzites and calc-silicates.
90
Metagreywakes appear as intercalations ranging from tens of centimetres to hundreds of
91
meters in thickness. Quartzites and calc-silicates form dismembered layers and enclaves of
3
92
mesosomes within the migmatites and they mark out a poorly preserved sedimentary layering
93
(Figure IIa). Amphibolitized dolerites and gabbros occur as dykes and sills and they are
94
coeval with the anatectic event as shown by magma mingling features between the mafic
95
intrusions and the granitic magmas. Isotopic data indicate Paleoproterozoic sources for both
96
sediments and magmas (Peucat et al., 1999 and references therein). According to Monnier et
97
al. (1996), mesosomes, neosomes and anatexites represent respectively 10%, 60% and 30% of
98
the rock types in the Pointe Géologie archipelago.
99
These migmatitic and anatectic formations underwent a polyphased tectonic and
100
metamorphic evolution with three phases of deformation leading to a composite sub vertical
101
regional S0-S1-S2 foliation (Monnier et al., 1996; Pelletier et al., 2002). Neosomes display two
102
successive parageneses related to both the S1 and S2 metamorphic foliations and bearing
103
comparable Sil+/-Kfs+/-Pl+/-Bt+/- Cd+/-Grt+/-Ilm+/-Qtz mineral assemblages. Moreover,
104
mineral chemistry of the S1 and S2 parageneses is similar and argues for a single continuous
105
LP-HT metamorphic event, the peak conditions of which reaching 750-700°C at 0.6-0.4 GPa.
106
Large amounts of anatectic melts have been produced all along the tectonic and metamorphic
107
evolution as suggested by the various habitus of the leucosomes. The oldest leucosomes
108
appear as mylonitic lenses isoclinally folded by the S1-S2 transposition, but most of them are
109
undeformed and concordant within the S2 foliation and the latest melts cross cut the S2 planes.
110
A retrograde imprint with Bt+/-Ms+/-And assemblages marks the later evolution to about
111
550-450°C and 0.5-0.4 GPa (Monnier et al., 1996; Pelletier et al., 2002; Gapais et al., 2008).
112
Bellair and Delbos (1962) assumed a 1.5 Ga age (Rb/Sr on micas) for the partial melting
113
event. More recently, new radiometric data (Peucat et al., 1999, U-Pb on zircons and
114
monazites, Sm-Nd and Rb-Sr on whole rocks and minerals) gave precise time constraints on
115
the geological evolution of the meta-volcano-sedimentary formations of Pointe Géologie.
116
Time of deposition is bracketed between the inherited zircon ages (1.72 -1.76 Ga) and the
117
peak metamorphism and partial melting age at 1.69 Ga. The later cooling path was
118
constrained by Nd garnet ages (650-600°C at 1.63-1.60 Ga) and Rb-Sr muscovite and biotite
119
ages (450-350°C at 1.5 Ga).
120
121
3. High temperature rocks types and regional structures
122
123
While the lithology is very homogeneous in the scale of the archipelago, field
124
relationships between different rock types are much more complex for the scale of outcrops.
125
Four main rocks types have been sampled for AMS measurements: migmatites including
4
126
leucosomes and melanosomes, coarse-grained pink granites, anatexites and strongly mylonitic
127
gneisses.
128
Migmatitic gneisses: They have been described previously as the K feldspar gneisses and
129
they are the dominant facies of the area, especially on the Pétrels Island (Dumont d’Urville
130
station, Figure I). We will discuss more specifically the space and time relationships between
131
leucosomes and melanosomes in the migmatitic gneisses. As partial melting developed all
132
along a polyphased tectonic and metamorphic event, leucosomes display various habitus
133
relative to the composite foliation, which is marked on the field by melanosomes and septas
134
of mesosomes. The melanosomes defined continuous layers in migmatites (Figure IIa):
135
(i)
The earliest leucosomes occur as mylonitized and elongated lenses, sometimes
136
affected by intrafolial folds within the foliation. They bear evidence of successive
137
segregation and deformation processes (Figure IIa).
138
(ii)
The most common leucosomes are undeformed and concordant within the
139
migmatites foliation and they may be later folded together with the foliation. They
140
represent syn to late-foliation melts (Figure IIb).
141
(iii)
The latest leucosomes form dykelets cross-cutting, and post-dating, the S1-2
142
foliation. They may be connected to the concordant leucosomes, tapping off and
143
transferring melts upwards (Figure IIa, c).
144
Coarse-grained pink granites: they appear as dykes of decimetric to decametric
145
thickness which cross cut the regional foliation (Figure IId). They are devoid of any visible
146
markers of orientation. Such dykes, collecting the dykelets, drain the more evolved melts
147
during the late- to post-tectonic stages of the partial melting and metamorphic event.
148
Anatexites: They outcrop widely on the W area of Pointe Géologie (e.g., Gouverneur,
149
Balance, Taureau and la Vierge islands, Figure VIII). They derived from a higher degree of
150
partial melting than migmatitic gneisses and show a mineral assemblage comparable to that of
151
leucosomes but more enriched in Qtz and depleted in Crd. Anatexites display more isotropic
152
textures but a flow fabric or a palimpsest structure of the parent gneisses is often marked by
153
some oriented rafts of migmatitic gneisses or schlierens of biotite and oxides (Figure IIe).
154
Mylonitic gneisses: they outcrop in a restricted and narrow zone on the western flank of
155
the Pétrels Island (Rocher Jakobsen) and some islands of Pointe Géologie, i.e. along a major
156
shear zone (sites II and V, Figure IX). They consist of Bt + Grt gneisses with only tiny Grt-
157
bearing leucosomes and are characterized by a vertical mylonitic foliation. They derive from
158
the K-feldspar migmatitic gneisses and then may be considered as late- to post-migmatisation
159
tectonites (Figure IIf).
5
160
As carefully described by Pelletier et al. (2002) and Gapais et al. (2008), the Pointe
161
Géologie archipelago show two types of structural domains: NS to N340° striking vertical
162
shear zones and domes both of them developing during the metamorphic and anatectic event.
163
Between the vertical zones, foliation attitudes point out dome-shaped structures marked by
164
flat lying to gently dipping fabrics. Dome displays a mineral lineation outlined by a preferred
165
orientation of sillimanite crystals while such a lineation is almost invisible in the vertical
166
zones. The transition from flat lying to vertical foliation zones may be sharp, with
167
development of vertical foliation planes cutting across the horizontal foliation of domes, or
168
gradual when folds, with horizontal axes, induce vertical planes. Such a transition from dome
169
to vertical shear zone never exceeds five hundred meters in thickness. In both cases, vertical
170
structures seem to be later than flat lying ones. According Gapais et al. (2008), the kinematic
171
markers (lineations, boudinage, ...) suggest that shear zones reflect a dextral transpressive
172
movement while flat areas are the result of shortening combined with stretching parallel to the
173
shear zones. Such patterns might be explained in the frame of a regional convergence of a hot
174
and weak juvenile crust leading to horizontal crustal flow and strain partitioning with
175
localized deformation along transpressional shear zones.
176
177
4. Sampling and analyses procedures
178
179
Because of outcrops restrictions (areas non accessible because of ice cover or of birds
180
protection) and high variability of the visible deformation, sampling was often made in sites
181
relatively close one to another in the main island (Pétrels Island). Other sites were scattered in
182
different smaller islands (Figure VIII). Most sampled sites correspond to a surface up to 50m2.
183
As a test, one site (site i, Figure IX) corresponds to a sampling made on a reduced surface
184
(about 0.5 m2) in gneisses showing visible foliation and lineation.
185
X large samples have been taken for petrographic and rock-magnetism studies. For
186
magnetic fabric study, cores were sampled with a gasoline-powered portable drill. All the
187
cores were oriented only with a sun compass (Figure IIf) because of the immediate proximity
188
of the southern magnetic pole. That introduced a strong limitation for sampling because of
189
climatic conditions. The sampling was made during three different summer fieldworks, in
190
1992, 2009 and 2011. As a whole, 819 rock cores were obtained in 33 sites, including 20 on
191
the Pétrels Island (Figure IX). When possible, different facies (from gneiss to anatexites) were
192
sampled on the same site. The cores were cut in the laboratory in specimen of standard
193
paleomagnetic size.
6
194
Magnetic susceptibility and its anisotropy have been measured using Kappabridges
195
KLY3 and MKF1 (AGICO, Brno). Thermomagnetic experiments were performed on
196
representative samples by heating in argon or in air using the CSL equipment and the CS2-3
197
oven (AGICO, Brno) associated with the Kappabridges. Hysteresis loops were obtained for
198
small cores (cylindrical samples of about 3 cm3) using a laboratory-made translation
199
inductometer within an electromagnet capable of reaching 1.6 T.
200
201
5. Magnetic Mineralogy
202
203
5.1. Thin sections study
204
As the magnetic characteristics of the rocks are tightly controlled by their
205
mineralogical content, the various assemblages recorded in the different rock types need to be
206
specified. This is particularly important in the case of migmatitic gneisses where partial
207
melting induces a redistribution of refractory minerals from the mesosomes to both the
208
melanosomes and leucosomes.
209
Magnetite and ilmenite are the most abundant oxide minerals. They form discrete grains or
210
clusters with direct contact between magnetite and ilmenite. The larger grains (millimetric in
211
size) are generally found in the melanosomes and therefore in the foliation and they appear as
212
elongated parallel to the lineation defined by the sillimanite when the latter is present (Figure
213
III). For ilmenite deformation also induces a structural axes related lattice preferred
214
orientation (LPO), which is characterized by the c-axes strongly concentrated close to the
215
normal to the foliation, and by the poles of (2-1-10) and (10-10) prism planes
216
concentrated close to the lineation (Figure IIIbis). Oxides are also present in the leucosomes
217
but the fabric (grains alignment) appears more scattered than in the melanosomes. At the
218
melanosome-leucosome border, elongated shape grains, may-be rotated, have been observed
219
perpendicular to the foliation. In the leucosomes oxide grains tend to be more discrete and
220
more rounded (Figure III).
221
Cordierite is also present in migmatitic gneisses and granitic dykes of Pointe Géologie.
222
This mineral that occurs mainly as grains of globular shape (up to a few millimetres in size) is
223
carefully watched because cordierite presents specific magnetic properties. This mineral
224
carries an inverse magnetic fabric (i.e. characterized by a switch between the principal
225
maximum K1 and minimum K3 axes of the AMS ellipsoid with respect to the grain shape
226
(Potter and Stephenson, 1988; Rochette et al., 1992)) that could lead to misinterpretation of
227
the measured magnetic fabrics. However the mean magnetic susceptibility (Km) for
7
228
magnetite, which is ubiquitous in studied thin sections, is significantly higher than for
229
cordierite (Km ratio > 1000). Therefore magnetite should be the main carrier of measured
230
AMS. For relatively large grains of magnetite, the K1 and K3 axes indicate the magnetic
231
lineation and magnetic foliation pole, respectively.
232
233
5.2. Rock-magnetism
234
The mean magnetic susceptibility Km (average 5.2 10-2) of the studied samples is
235
different according to the studied fraction of the migmatites (Figure IV). The gneisses with
236
subordinate melted parts and dominant melanosome fraction have very high susceptibility
237
(average?). Km values for dominant granitic fraction are lower (average?). Migmatite samples
238
including both leucosome and melanosome fractions present intermediate Km values. This
239
variability in leucosome versus melanosome content of gneissic specimens largely explains
240
the relatively large distribution of susceptibilities (Figure IV). The anatexites samples show
241
the highest variation in Km values (from X to Y).
242
All thermomagnetic KT curves (susceptibility K as a function of temperature T) show
243
during heating a sharp susceptibility decrease around 580°C (Figure V), clearly indicating the
244
presence of magnetite. In addition low temperature experiments confirm the presence of this
245
mineral by a large susceptibility increase around -160°C (Figure V). This variation
246
corresponds to the Verwey transition that is characteristic of pure magnetite. Few
247
thermomagnetic curves are perfectly reversible. Susceptibility values from the cooling curve
248
are mostly a little bit higher than the initial values, in air as in argon atmosphere (but
249
sometimes much higher as for the sample AP35). Partial cooling loops during heating
250
evidenced a weak mineral alteration due to heating around 300 to 400°C that probably reflects
251
inversion of a weak amount of maghemite in hematite. New magnetite is also often formed at
252
the highest temperatures.
253
Hysteresis loops show generally a full saturation of the magnetization for moderate
254
field (Figure VI). This saturation was partial only for the sample AP35, evidencing the
255
presence of another component of high coercivity. This component is probably hematite. In
256
most cases, ferrimagnetic minerals have a largely dominant effect for the susceptibility, but
257
two of the studied granite samples highlight also a strong contribution of the paramagnetic or
258
even diamagnetic minerals. Day plot (Day et al., 1977) for samples with mainly magnetite
259
clearly evidences a large size of the magnetite particles (Figure VII).
8
260
Main magnetic mineral is therefore magnetite with a large size. Obtained AMS data
261
are then directly usable for a structural analysis (no inverse magnetic fabric due to small
262
Single Domain particles (Potter and Stephenson, 1988).
263
264
6 - Magnetic fabric
265
266
The AMS ellipsoid yields the three principal susceptibility axes defined as
267
K1  K2  K3. K1 is the magnetic lineation and K3 is perpendicular to the magnetic foliation.
268
The Jelinek (1981) parameters P' and T were used to describe the intensity and the shape of
269
the magnetic fabric, respectively. Data for a group of samples were analysed using normalized
270
tensor variability (Hext, 1963; Jelinek, 1978) statistics.
271
6.1 – Structures from AMS
272
Data for gneiss and for granite are presented separately (often "gneisses" included
273
minor melted parts and "granitic" samples small amounts of melanosome minerals and this
274
separation has been made then considering the dominant presence of one of these facies).
275
Moreover, from the regional structural field pattern, two distinct zones have been
276
distinguished for the AMS analyse: the sub-vertical zones and the dome shaped-zones (Figure
277
VIII and IX).
278
6.1.1. In migmatitic gneisses of sub-vertical shear zones (sites T, R, D, O, XIII, A of
279
Pétrels Island, Figure IX, melanosome layers show generally both well clustered magnetic
280
foliation pole (K3) and magnetic lineation (K1). The magnetic foliation is dominantly
281
subvertical N-S in accord with the observed field foliation and the magnetic lineation presents
282
a weak plunge and N-S trending in agreement with the orientation of mineral lineation
283
outlined in few places by a preferred alignment of sillimanite crystals. That is in particular the
284
case in the site test i (Figure IX) where foliation and lineation are clearly visible. The granitic
285
rich layers (site XIII) display more diffuse fabric, in particular the K1 axes that tend to form a
286
girdle perpendicular to K3. Well-individualized leucosomes (e.g. Figure IIC) show a different
287
magnetic fabric signature than melanosome layers characterized by strongly inclined to sub-
288
vertical magnetic lineation (site XIII
289
than the K1 axes. They are in the horizontal plane with two dominant directions: N-S and E-
290
W.
291
292
. The K3 axes are more scattered
In large coarse-grained pink granite dykes (sites A, E, D
IXthe K3
axes are relatively well clustered defining a subvertical magnetic of strike close to that of
9
a divergent strike (sites A). The
293
neighbouring gneisses (sites E, D
294
magnetic lineation pattern is characterized by the verticalization of K1-axes in comparison to
295
magnetic lineation measured in the melanosome layers of neighbouring gneisses. They form
296
girdle perpendicular to K3 (site E or strongly inclined to sub-vertical local maximum
297
(sitesD A

298
, Taureau, la Vierge, Balance Islands (sites
299
VIIXV, XIV XVIII, XIXFigure VIII) and South of Pétrels Island (site
300
XX
301
Gouverneur anatexite (sites VII, XV
302
lies in the horizontal plane, relatively scatted or forms clusters at 90 degrees from each other
303
with the N-S and E-W directions that prevails for south Gouverneur Islands (site XV).
304
Sampling sites XIV, XVIII (Figure VIII) show heterogeneities in the partial melting degree
305
that is expressed in the field by more or less regularly presence of migmatitic gneiss rafts in
306
the anatexites. For site XVIII
307
gneisses show distinct magnetic axes orientations characterized by K1 preferentially sub-
308
vertical and K3 horizontal N-S (i.e., magnetic foliation vertical E-W) in anatexites while K1
309
is horizontal to strongly plunging N-S and K3 is horizontal E-W (i.e., magnetic foliation
310
vertical N-S) in neighbouring gneisses. The N-S orientation of magnetic foliation in gneisses
311
fit quite well with the regional vertical shear zone structures. In opposite, the E-W vertical
312
magnetic foliation that appears in the larger dykes and anatexites is less well-constrained by
313
large-scale
314
XIV T
315
and the migmatitic gneisses was difficult to perform. That explains the relatively wide
316
confidence zones at 95% for K1 and K3. However the magnetic fabric is mainly oriented
317
vertical E-W for the foliation and strongly-inclined (westward) for the lineation.
318
K1max tends to have strong inclination or to be sub-vertical. For
field
structures
 anatexites and migmatitic
(Figure
VIII).
For
site
tinction on the field between the anatexites
In mylonitic gneisses of Jakobsen (sites II, V,
) the magnetic foliation
319
(K1-K2) is dominantly subvertical N-S in accord with the observed field structures (Figure
320
IIf). On the contrary, the AMS measurement reveals uncommon highly dipping or sub-
321
vertical lineation.
322
323
6.1.2. In the migmatitic gneisses of the dome shaped-zones (sites J, VI, G, K, H, S,
324
XVI, XVII, Figure VIII; sites C, IV, Q, III, VIII, I, B, i, XII, F,
) magnetic
325
fabric display a stronger variability of K1 and K3-axes orientations from one site to another
10
326
one in comparison to structured vertical shear zones. A relatively strong dip (about 45°) for
327
the magnetic foliation is frequently observed, in particular for the sites located in the flanks of
328
domes (sites H, K, Figure VIII; sites IV, Q, III, Figure IX) for the eastern flanks of the
329
Pétrels-Curies Islands dome). The magnetic foliations are generally in agreement with the
330
field measured strike and dip of melanosome layers. The magnetic lineations show gently dip
331
and dominant NNW strike. The innermost zones of domes (sites XVI, XVII, Figure VIII; sites
332
VIII, I, B, Figure IX) show also a strong variability of AMS structure patterns correlated to
333
the complexity of migmatitic structures observed at the domes apex. In these areas are
334
observed the less well-defined magnetic directions (with the widest confidence zones at 95%:
335
site XVI, Figure VIII; site I, Figure IX). Sampling in a granitic dykelet of centimetric width
336
crosscutting flat-lying gneissic foliation was performed in site XVII (Figure VIII). The
337
related AMS pattern is characterized by vertical N-S magnetic foliation and sub-vertical
338
magnetic lineation and then differs from that of gneiss.
339
Broad dyke of coarse grained pink granites sampling in site F an XII (Figure IX) are
340
characterized as for the granitic sites located in vertical shear zone by verticalization of
341
magnetic lineation in comparison to the magnetic lineation measured in the melanosomes
342
layers of gneisses The magnetic foliation in the dykes is either the same direction as that
343
measured in the gneiss (site F) or discordant (site XII).
344
345
6.2 – AMS parameters (P’ and T)
346
The Figure X presents the characteristics of the susceptibility ellipsoid. The degree of
347
anisotropy P’ is very high in the gneisses (mean 1.479; minimum 1.041; maximum 1.903). It
348
is lower in granites (mean 1.222; minimum 1.040; maximum 1.413). The shape of the
349
susceptibility ellipsoid is dominantly oblate, in gneisses (for T value: mean 0.30;
350
minimum -0.08; maximum 0.66) as in granites (mean 0.25; minimum -0.10; maximum 0.62).
351
In granites, a clear relationship (Figure IV, R2=0.70) between P’ (intensity of the fabric) and
352
Km (related to the concentration in ferrimagnetic minerals) highlights the dominant effect of
353
the magnetite in the AMS (Henry et al., 2004). Another relationship (R2=0.54) between T and
354
P’ in granites suggests that the most prolate fabrics correspond to the weakest anisotropy
355
Figure X. In addition, for granite, the anatexites show the highest variability in P’ and T
356
values, including the extreme values Figure X.
357
11
358
7 – Discussion
359
7.1 Mean susceptibility (Km) and anisotropy (P’) signature
360
The amount of magnetite is clearly different in gneisses and in granites (leucosomes
361
and dyke), as shown by the susceptibility values (Figure IV). This variation of magnetic
362
susceptibility between melanocratic and leucocratic layers is mainly due to the variations in
363
magnetite content. Magnetic studies suggest also mineralogical changes affecting magnetite
364
with formation of ferromagnetic oxides of lower susceptibility like maghemite or hematite.
365
Concerning this mineralogical change, gneiss sample AP35 is the only one with hematite; it
366
also presents the strongest formation of a lot of new magnetite at high temperature
367
(susceptibility becomes 14 times higher on the cooling curve). It is therefore probably due to a
368
strong local mineralogical alteration (interaction with deep fluids, weathering?) and remains
369
an exception. In all the other gneissic samples, no or few maghemite alteration during
370
thermomagnetic experiments has been observed, while such alteration is more important in
371
the granitic ones. This mechanism therefore likely had a significant effect for the
372
susceptibility decrease in granites related to melting.
373
The intensity of magnetic anisotropy, illustrated by the P’ values, is lower in granites
374
than in gneisses (Figure X). In granites, it presents a relationship with the Km values,
375
suggesting that this difference could be explained by the lower amount of magnetite in
376
granites, AMS then resulting mainly from strong magnetite anisotropy and weak anisotropy
377
of a low susceptibility component. Owing to its Km values, this last component is the
378
paramagnetic minerals and not the maghemite. The observed inverse correlation between T
379
and P’ in granites (Figure X) shows that the AMS of this second component is less oblate than
380
the magnetite AMS.
381
Inversely to granitic magmas constituting leucosomes and dykes, a wide range in Km
382
values, according to the samples within a same site, is observed in anatexites. It is due to the
383
presence of more refractory slice of gneisses with high concentration in oxides grains.
384
385
7.2 AMS in gneisses and granites
386
In addition to the Km and P’ values, the main differences in magnetic fabric between
387
melanosome layers of gneisses and granitic magmas concern the orientation of the K1 axis
388
(Figures VIII and IX). K1 axis is dominantly close to the horizontal in melanosomes and close
389
to vertical in granites. In some granitic sites (e. g., sites E, XII, Figure IX), its orientation is
390
relatively scattered within a vertical plane, probably because of the presence of melanosome
391
fragments within the granite. Other difference in orientation of susceptibility axes concerns
12
392
K3. Whereas in melanosome the K3 orientation fits to the field structures, it appears in granite
393
a more complex magnetic orientation pattern with two dominant orientations. The first one
394
matches to the gneissic structures (for example, the magnetic foliation is vertical N-S in the
395
shear zone area) and the second is at 90 degree of the first one. In some sites the two
396
clustering are present (e.g., site XV Figure VIII and site XIII, Figure IX). The first
397
orientation is associated to the strike-slip deformation, which is at the origin of the large-scale
398
sub-vertical corridors. The cause of the second orientation is less straightforward. In Pétrels
399
Island, the migmatitic layering is locally cut by vertical large granitic dykes of NE—SW to E-
400
W direction (Monnier, 1995). The E-W tending magmatic direction could be then related with
401
the implementation of dykes, which are consistent with a component of NNW-SSE to N-S
402
dextral strike-slip (Gapais et al., 2008). In the western area (Gouverneur island area, Figure
403
VIII) where pieces of melanocrate restites are still packed in anatexites, E-W magnetite could
404
correspond to inheritance of old tectonic foliations.
405
The gneiss sites II and V, on the south-western border of the Pétrels Island (Figure IX)
406
also present mainly subvertical K1 axes. However, contrarily to the granitic leucosome, dyke
407
and anatexite that have not recorded pervasive solid-state deformation, the gneisses of sites II
408
and V are intensely deformed. Thus, AMS fabric is interpreted as resulting of high-strained,
409
late- to post-migmatisation mylonites. The strong K3 clustering indicating a N-S vertical
410
magnetic foliation and the highly tilted to sub-vertical magnetic lineation could be related to
411
deformation in transpressionnal regime that is localised in the N-S vertical shear zones.
412
Thin section analyses show a change of orientation of magnetite grains from
413
melanosome to leucosome. In melanosome, the oxide grains are in the foliation elongated
414
parallel to the lineation marked by the sillimanite alignment. They are interpreted as resulting
415
from solid-state deformation that occurred during the Paleoproterozoic metamorphic event
416
(1.7 Ga ago). In addition, LPO of ilmenite, which strongly correlated with the tectonic fabric,
417
suggest dislocation glide on (0001) as main deformation mechanism. The basal slip for
418
hematite – ilmenite mineral could reflect high-temperature solid-state deformation (e.g.,
419
Bascou et al., 2002; Siemes et al., 2003). Migmatitisation processes are also associated with
420
segregation and transportation of melt that result in leucosome, ping granitic dyke and
421
anatexite in the area of higher degree of partial melting. As discussed in Vigneresse et al.
422
(1996), the increase of melt fraction produce changes in rheological behaviour. Plastic
423
deformation evidenced by elongation of oxide minerals, which are abundant in the
424
melanosome and magma flow with transport of relatively isolated solid particles in
425
leucosomes, dykes and anatexites.
13
426
Such types of deformation succession involving magnetite grains deformed under
427
solid-state conditions and, during melting, behaving like passive rigid particles has been
428
observed in various migmatitic terrains (e.g., Barbey et al., 1990; Ferré et al., 2003). In Pointe
429
Géologie migmatites, this generates distinctive AMS signatures. The clearest difference
430
concerns the K1 orientation, which tends to a vertical plunge during segregation and
431
circulation of felsic melt. This is related to gravity-induced upwelling of magmas. K3
432
orientation in granitic layer appears more sensitive to the degree of partial melting and the
433
structuration of neighbouring rock. When the melting started, melt segregation remained
434
strongly constrained by unfused material. Felsic melt formed dykelets and connected lenses
435
concordant with the foliation of gneisses. For higher degree of melting, the melt was injected
436
as concordant or cross-cutting large veins (Figure IIc,d) explaining the various orientations of
437
the magnetic foliation. The dyke emplacement corresponds to areas of lower stress, generally
438
close to the 3 direction (e.g., Cooper, 1990; Hand and Dirks, 1992).
439
440
441
7.3 Methodological and regional-scale implications
442
All these observations evidence that, for migmatites, the AMS study is clearly not a
443
routine method, the single possible approach having to be based on the obtaining of numerous
444
data from rocks showing various degree of melting. For low degree of partial melting the
445
leucocratic layers remain strongly constrained by the structure defined by the unfusible
446
minerals. The AMS principal axes are representative of the local and regional finite strain
447
axes. This observation is conform to those of Ferré et al. (2003) for migmatites of the
448
Superior Province in Minnesota. For more pronounced partial melting degree, well-
449
individualized leucocratic layers can developed specific AMS patterns associated to flow
450
regime. In Pointe Géologie migmatites the difference with the local and regional finite strain
451
axes is marked by the direction of K1-axes. Probability to measure composite AMS is
452
increased and the difficulty of AMS analyses is then to isolate pre-melting and synmelting
453
magnetic fabrics. The anatexites show the strongest variations in AMS principal directions
454
and parameters due to the inhomogeneity of these formations, which can preserve more
455
refractory pieces.
456
The structural pattern of Pointe Géologie is constituted by two superimposed
457
foliations, one flat-lying and one sub-vertical, both associated with partial melting (Pelletier et
458
al., 2002; Ménot et al., 2007). Mineral lineations are not easy to observe within the foliation,
459
in particular for sub-vertical shear zones. These structural field evidences, coupled with the
14
460
metamorphic pattern and ductile deformation numerical simulations, allowed to propose a
461
model in which hot and weak lithosphere converge during Paleoproterozoic toward a cooler
462
craton (Archean in age). In this model, deformation combines transpression and horizontal
463
flow and/or lateral escape parallel to the Archean basement (Gapais et al., 2008).
464
AMS structural pattern determination provides complementary structural information
465
such as lineations. They are mainly sub-horizontal in sub-vertical foliated gneisses. It
466
underlines the importance of the strike-slip component during the deformation. In mylonitic
467
gneisses that are considered as late- to post-migmatisation tectonites appear lineations with
468
stronger plunge. This is attributed to the effect of transpressionnal regime, which combinates
469
strike-slip and across-strike shortening (ref? voir Tikoff). In migmatite with flat-lying
470
foliation, lineation has moderate plunge, which is conform to field observations.
471
Increase of partial melting degree produce decrease of the grains cohesion and the melt
472
escapes of the frame constituted of solid-state deformed minerals. Thus the former tectonic
473
structures will tend to be erased. Segregated melt will behave as viscous flow and movement
474
are gravity driven forces. Well-defined leucosomes, dykes and anatexites will develop new
475
magnetic fabric corresponding to the transport of the magma upwards. Measurements of AMS
476
then allow to better map the mass transfers associated to the peak temperature and intensive
477
melting and then specify the general framework of a warm and buoyant lithosphere
478
converging toward a cooler craton as proposed for Pointe Géologie.
479
480
8 – Conclusion
481
Pointe Géologie archipelago in Terre Adélie represents a deep and hot crust affected
482
by intensive anatectic event during Paleoproterozoic times, 1.69 Ga ago. This generated
483
different HT rocks types such as: migmatites including leucosomes and melanosomes, coarse-
484
grained pink granites dyke and anatexites. The partial melting affects indifferently the
485
regional structures characterized by the juxtaposition of flat-lying and sub-vertical
486
deformation zones.
487
Magnetic mineralogy study shows that oxide minerals are mainly large grains of
488
magnetite. Ilmenite is also observed constituting cluster with magnetite, which is particularly
489
abundant in melanosome layers. The grains are intensely solid-state deformed and aligned
490
parallel to the lineation marked by sillimanite. Oxides appear more scattered in the
491
leucosomes than in the melanosomes. Their rounded shape suggests rotation of the former
492
elongated grain during melt segregation.
15
493
Melanocratic layers on the one hand and felsic leucosomes, dykes and anatexites on
494
the other hand develop contrasted AMS signatures. The mean magnetic susceptibility (Km)
495
and the anisotropy degree (P’) are high in the gneisses and lower in granitic samples.
496
Orientation of principal susceptibility axes also differs. In gneisses, magnetic foliations are
497
coherent with the local and regional structural pattern. Nevertheless AMS technique provides
498
complementary information on the orientation of lineations, which are difficult to observe in
499
the field. They are subhorizontal in shear zone, except in vertical mylonitic zone where high
500
plunge evidences the transpressionnal regime of the deformation. In well-defined leucosomes,
501
granitic dykes and anatexites, the magnetic ellipsoid is characterized by verticalization of the
502
magnetic lineation. This change of orientation is associated to a rheological contrast between
503
the solid-state deformation suffered by the oxide grains and their reorientation in a viscous
504
flow during the aggregation of felsic melt.
505
Magnetic structure pattern do not disagree with model of hot and weak lithosphere
506
converging toward a cooler Archean craton proposed by Gapais et al. (2008) from
507
metamorphic and field structures analysis, and from numerical modelling (Duclaux et al.,
508
2007). The auteurs focussed particularly on structural features compatible with a horizontal,
509
orogen-parallel, subconstrictional flow that could result from oblique convergence and/or
510
from particular boundary conditions facilitating lateral escape. The magnetic structures map
511
allows to highlight the role of the gravity-induced upwelling of a crust undergoing partial
512
melting.
513
From methodological point of view, this work emphases on the importance of the
514
petrological characterization of sites being selectioned for AMS measurements in order to
515
avoid pre-melting and synmelting composite magnetic fabrics. Finally, it underline the
516
difficulty of AMS analyse in anatexite, which presents the larger variability for AMS
517
signature in Pointe Géologie, due to the heterogeneity of the magma and the preservation of
518
refractory tectonite slices.
519
520
Acknowledgements
521
We are very grateful to the "Institut polaire français Paul Emile Victor" for support of
522
the program ArLiTA (Architecture de la Lithosphere de Terre Adélie) and to the Japanese
523
"National Institute of Polar Research" for participation to the international exchange program
524
under the Antarctica Treaty and for invitation of BH in Tokyo. LPO were carried from EBSD
525
measurements using a SEM-JSM 5600 at the university of Montpellier (France) and ilmenite
16
526
pole figures were generate from software developed by D. Mainprice (Geosciences
527
Montpellier).
528
529
References
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
Barbey, P., Macaudière, J., Nzenti, J.P., 1990. High-pressure deshydratation melting of
metapelites. Evidences from the migmatites of Yaoundé (Cameroon). J. Petrol. 31, 401427.
Bascou, J., Raposo, M.I.B., Vauchez, A., Egydio-Silva, M., 2002. Titanohematite lattice
preferred orientation and magnetic anisotropy in high-temperature mylonites. Earth
Planet. Sci. Lett. 198, 77-92.
Bellair, P., 1961a. Pétrographie du socle cristallin de la Terre Adélie. C. R. Acad. Sci.,
Paris D252, 3296-3298.
Bellair, P., 1961b. Sur les formations anciennes de l'archipel de Pointe Géologie (Terre
Adélie). C. R. Acad. Sci., Paris D252, 3087-3089.
Bellair, P., Delbos, L., 1962. Age absolu de la dernière granitisation en Terre Adélie. C.
R. Acad. Sci., Paris D254, 1465-1466.
Borradaile, G.J., Henry, B., 1997. Tectonic applications of magnetic susceptibility and
its anisotropy. Earth Sci. Rev. 42, 49-93.
Borradaile, G.J., Jackson, M., 2004. Anisotropy of magnetic susceptibility (AMS),
magnetic petrofabrics of deformed rocks, Magnetic fabrics. The Geological Society
London, special publication, pp. 299-360.
Charles, N., Faure, M., Chen, Y., 2009. The Montagne Noire migmatitic dome
emplacement (French Massif Central): new insights from petrofabric and AMS studies.
J. Struct. Geol. 31, 1423-1440.
Cooper, R.F., 1990. Differential stress-induced melt migration: an experimental
approach. J. Geophys. Res. 95, 6979-6992.
Day, R., Fuller, M., Schmidt, V.A., 1977. Hysteresis properties of titanomagnetites:
grain size and compositional dependence. Phys. Earth Planet. Int. 13, 260-267.
Duclaux, G., Rey, P., Guillot, S., Ménot, R.P., 2007. Orogen-parallel flow during
continental convergence: Numerical experiments and Archean field examples. Geology
35, 715-718.
Fanning, C.M., Daly, S.J., Bennett, V.C., Ménot, R.P., Peucat, J.J., Oliver, R.L.,
Monnier, O., 1995. The "Mawson Block" : once contiguous Archean to Proterozoic
crust in the East Antarctic Shield and Gawler Craton, Australia, VII International
Symposium on Antarctic Earth Sciences, Spec. Vol. Terra Antartica, Siena, Italy.
Fanning, C.M., Ménot, R.P., Peucat, J.J., Pelletier, A., 2002. A closer examination of the
direct links between Southern Australia and Terre Adélie and Georges V land,
Antarctica, 16th Australian Geolical Congress, AGC, Adélaïde, Australia.
Ferré, E., Teyssier, C., Jackson, M., Thill, J.W., Rainey, E., 2003. Magnetic
Susceptibility Anisotropy: A new petrofabric tool in migmatites. J. Geophys. Res.
108(B2), 20086-20100.
Gapais, D., Pelletier, A., Ménot, R.P., Peucat, J.J., 2008. Paleoproterozoic tectonics in
the Terre Adélie Craton (East Antarctica). Precambrian Res. 162, 531-539.
Graham, J.W., 1954. Magnetic susceptibility anisotropy: an unexploited element of
petrofabric. Geol. Soc. Am. Bull. 65, 1257-1258.
17
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
Hand, M., Dirks, P.H.G.M., 1992. The influence of déformation of axial planar
leucosomes and the segregation of small melt bodies within the Napperby gneiss, Central
Australia. J. Struct. Geol. 14, 591-604.
Hasalovà, P., Schulmann, K., Lexa, O., Štipskà, P., Hrouda, F., Ulrich, S., Haloda, J.,
Týcovà, P., 2008. Origin of migmatites by deformation-enhanced melt infiltration of
orthogneiss: a new model based on quantitative microstructural analysis. J.
metamorphic Geol. 26, 29-53.
Henry, B., Merabet, N., Derder, M.E.M., Bayou, B., 2004. Chemical remagnetizations in
the Illizi basin (Saharan craton, Algeria) and their acquisition process. Geophys. J. Int.
156, 200-212.
Hext, G., 1963. The estimation of second-order tensors, with related tests and designs.
Biometrika 50, 353.
Jelinek, V., 1978. Statistical processing of anisotropy of magnetic susceptibility
measured on groups of specimens. Studia Geoph. Geod. 22, 50-62.
Jelinek, V., 1981. Characterization of the magnetic fabrics of rocks. Tectonophysics 79,
63-67.
Kratinová, Z., Schulmann, K., Edel, J.-B., Tabaud, A.-S., 2012. AMS record of brittle
dilation, viscous-stretching and gravity-driven magma ascent in area of magma-rich
crustal extension (Vosges Mts., NE France). Int J Earth Sci 101, 803-817.
Kruckenberg, S.C., Ferré, E.C., Teyssier, C., Vanderhaeghe, O., Whitney, D.L., Seaton,
N.C.A., Skord, J.A., 2010. Viscoplastic flow in migmatites deduced from fabric
anisotropy: An example from the Naxos dome, Greece. J. Geophys. Res. 115, B09401.
Ménot, R.P., Duclaux, G., Peucat, J.J., Rolland, Y., Guillot, S., Fanning, C.M., Bascou,
J., Gapais, D., Pêcher, A., 2007. Geology of the Terre Adélie Craton (135 - 146° E), 10th
ISAES. USGS, Santa Barbara.
Monnier, O., 1995. Le socle protérozoïque de Terre Adélie (Antarctique Est): son
évolution tectonométamorphique et sa place dans les reconstitutions du protogondwana. Jean Monnet, Saint-Etienne.
Monnier, O., Menot, R.P., Peucat, J.J., Fanning, M., Giret, A., 1996. Actualisation des
données géologiques sur Terre Adélie (Antarctique est): mise en évidence d'un collage
tectonique au Protérozoïque. C. R. Acad. Sci., Paris 322, 55-62.
Pelletier, A., Gapais, D., Ménot, R.P., Peucat, J.J., 2002. Tectonique transpressive en
Terre Adélie au Paléoprotérozoïque (East Antarctica). C. R. Acad. Sci., Paris 334, 505511.
Peucat, J.J., Ménot, R.P., Fanning, C.M., 1999. The Terre Adelie basement in the EastAntarctica Shield: geological and isotopic evidence for a major 1.7 Ga thermal event;
comparison with the Gawler Craton in south Australia. Precambrian Res. 94, 205-224.
Potter, D.K., Stephenson, A., 1988. Single domain particles in rocks and magnetic fabric
analysis. Geophys. Res. Lett. 15, 1097-1100.
Rochette, P., Jackson, M., Aubourg, C., 1992. Rock magnetism and the interpretation of
anisotropy of magnetic susceptibility. Rev. Geophys. 30, 209-226.
Schulmann, K., Edel, J.B., Hasalová, P., Cosgrove, J., Jések, J., Lexa, O., 2009. Influence
of melt induced mechanical anisotropy on the magnetic fabrics and rheology of
deforming migmatites, Central Vosges, France. J. Struct. Geol. 31, 1223-1237.
Siemes, H., Klingenberg, B., Rybacki, E., Naumann, M., Schäfer, W., Jansen, E.,
Rosière, C.A., 2003. Texture, microstructure, and strength of hematite ores
experimentally deformed in the temperature range 600–1100 °C and at strain rates
between 10-4 and 10-6 s-1. J. Struct. Geol. 25, 12371-11391.
Tarling, D.H., Hrouda, F., 1993. The Magnetic Anisotropy of Rocks, First edition 1993
ed. Chapman and Hall, London.
18
622
623
624
625
626
Vigneresse, J.L., Barbey, P., Cuney, M., 1996. Rheological transitions during partial
melting and crystallisation with application to felsic magma segregation and transfer. J.
Petrol. 57, 1579-1600.
19