
Significance of the Nova Brasilândia metasedimentary belt

in western Brazil: Redefining the Mesoproterozoic

boundary of the Amazon craton

Eric Tohver,1,2 Ben van der Pluijm,1 Klaus Mezger,3 Eric Essene,1 Jaime Scandolara,4

and Gilmar Rizzotto4

Received 18 July 2003; revised 6 July 2004; accepted 16 September 2004; published 7 December 2004.

[1] The Nova Brasilândia metasedimentary belt
(NBMB) of western Brazil marks a fundamental
crustal boundary in the Amazon craton. The
metasedimentary rocks of the NBMB (calc-silicates,
metapelites, quartzites, metabasites) contrast strongly
with the older, polycyclic granitoid rocks of the
adjacent Amazon craton. Aeromagnetic anomalies
indicate that the belt is continuous for at least
1000 km in an E-W direction, although the
easternmost extent of the NBMB is covered by
the Cretaceous sediments of the Parecis Formation.
Additional geologic evidence suggests that the belt
extends along an E-W trend for �2000 km. The
northern portion of the NBMB preserves vestiges of
an early high pressure-temperature (P-T) assemblage
(kyanite + staurolite) overprinted by sillimanite during
prograde metamorphism. A higher metamorphic grade
is observed in the southern portion of the belt,
with peak conditions calculated to be 800 MPa and
800�C for granulitic assemblages. The combined P-T
path demonstrates that the competing processes
of imbrication (northern domain) and magma
generation (southern domain) are responsible for
regional metamorphism and crustal thickening.
Cooling from peak metamorphic conditions is
recorded by U-Pb monazite ages of 1090 Ma and
titanite ages of �1060 Ma. Integrated cooling rates
of 2�–3�C/Myr from regional metamorphism are
calculated from these U/Pb ages combined with
40Ar/39Ar ages of hornblende (�970 Ma) and biotite
(�910 Ma). The NBMB marks the Mesoproterozoic
limit of the SW Amazon craton. The discordance
of the NBMB to the NNW structural trend of the
younger Aguapeı́ belt (200 km SE of NBMB),
together with marked differences between the two

belts in sedimentary environment, metamorphic grade,
and timing of deformation, signify that these two belts
are not geologically continuous. The ‘‘Grenvillian’’
deformation recorded by the NBMB belt marks the
final docking of the Amazon craton and Paragua
craton within the Rodinia framework. The Aguapeı́
belt, in contrast, seems to record only limited
deformation internal to the Paragua craton. INDEX

TERMS: 8102 Tectonophysics: Continental contractional orogenic

belts; 3660 Mineralogy and Petrology: Metamorphic petrology;

1035 Geochemistry: Geochronology; 8025 Structural Geology:

Mesoscopic fabrics; 9360 Information Related to Geographic

Region: South America; KEYWORDS: Rodinia, Grenville mobile

belt, Amazon craton, Paragua craton, geochronology, P-T path.

Citation: Tohver, E., B. van der Pluijm, K. Mezger, E. Essene,

J. Scandolara, and G. Rizzotto (2004), Significance of the Nova

Brasilândia metasedimentary belt in western Brazil: Redefining

the Mesoproterozoic boundary of the Amazon craton, Tectonics,

23, TC6004, doi:10.1029/2003TC001563.

1. Introduction

[2] Many recent plate reconstructions for the Mesoprote-
rozoic postulate that the Grenville orogeny represents world-
wide events leading to the assembly of the supercontinent
Rodinia [Dalziel, 1991; Hoffman, 1991]. North America
holds a keystone position in Rodinia based on the Neo-
proterozoic rift basins that circumscribe Laurentia [Bond et
al., 1984], with most reconstructions placing the southwest-
ern margin of Amazonia in proximity to some portion of
eastern Laurentia. The Grenvillian history of the eastern
Laurentian margin has been relatively well documented
along the �3000 km length of this ancient mobile belt,
which stretches from the southernmost portion of cratonic
Laurentia typified by the Llano Uplift region of central Texas
[Mosher, 1998] to the northeastern exposures of Labrador
[Krogh, 1994; Davidson, 1998]. In contrast, present knowl-
edge of the Grenvillian history of the Amazon craton is less
complete, making exposed portions of the belts all the more
interesting from the standpoint of Rodinia paleogeography.
[3] Three positions have been proposed for the Meso-

proterozoic paleogeography of the Amazonia-Laurentia
connection within Rodinia, based on different lines of
evidence. Hoffman [1991] and Sadowski and Bettencourt
[1996] used broad geological similarities, such as igneous
age provinces or metamorphic ages, to pair Amazonia with
the central Grenville Province of Ontario and New York at a

TECTONICS, VOL. 23, TC6004, doi:10.1029/2003TC001563, 2004

1Department of Geological Sciences, University of Michigan, Ann
Arbor, Michigan, USA.

2Now at the Instituto de Geociências (GMG), Universidade de São
Paulo, Cidade Universitária, São Paulo, SP Brazil.

3Institut für Mineralogie, Universität Münster, Münster, Germany.
4Companhia de Pesquisas de Recursos Minerais (CPRM), Porto Velho,

RO Brazil.

Copyright 2004 by the American Geophysical Union.
0278-7407/04/2003TC001563$12.00

TC6004 1 of 20


later interval, circa 1.0 Ga. Dalziel [1991] chose a more
northerly position for Amazonia alongside Newfoundland
and Labrador, based largely on matching the Iapetan rifting
histories of Laurentia with the Precordilleran margin of
South America [Bond et al., 1984], which is now recog-
nized to constitute a separate terrane, exotic to South
America [Thomas and Astini, 1996]. New paleomagnetic
data for the early ‘‘Grenvillian’’ interval (circa 1.2 Ga) of
the Amazon craton supports a collision with the southern-
most portion of Laurentia, i.e., the Llano region of central
Texas [Tohver et al., 2002]. Observations of large-scale
strike-slip deformation of the Amazon craton basement
rocks, together with observations from Laurentia, suggest
large-scale displacement between Amazon and Laurentia,
possibly reconciling the Hoffman [1991] and Tohver et al.
[2002] paleographic models. According to this transpressive
model of Grenville orogenesis, the evolution of Rodinia was
marked by both thrust faulting and strike-slip tectonics
along the greater Grenville collisional zone, comprised of
both its North and South American equivalents.
[4] One of the principal tasks in placing Amazonia in a

Rodinian reference frame is identifying the craton’s principal
structures, particularly those that are relevant to its Meso/
Neoproterozoic history. The classical work of Almeida and
Hasui [1984] defines the Amazon craton at its southern and
eastern margins by the presence of Brasiliano/Panafrican
belts (double black line, Figure 1, bottom inset). However,
the emphasis on structures related to Gondwanan history
distracts from the somewhat older signs of a previous
supercontinent, thus obscuring the role of Amazonia in the
constructive phase of Rodinian history. We contend that the
Meso/Neoproterozoic cratonic margin of Amazonia is such a
structure overlooked by the traditional model. In this con-
tribution, new geochronological, petrological, and structural
data from the Nova Brasilândia Metasedimentary belt
(NBMB) are presented as evidence of high-grade deforma-
tion under transpressive conditions during late Mesoprotero-
zoic times. Previously published aeromagnetic data from this
belt suggest that lithological and structural trends in the
NBMB continue from the study region �1000 km to the
eastern limit of the Amazon craton, the Brasiliano-aged
Paraguai belt. This eastern portion of the belt is entirely
covered by the Mesozoic sediments of the Parecis Formation
[Barros et al., 1978]. Approximately 750 km to the west,
K-Ar geochronology on borehole samples of the metasedi-
mentary basement underlying the Phanerozoic sediments in
the westernmost Amazon basin on the Brazil/Peru border
yields ages (circa 1.0 Ga) similar to those observed in our
study region [Amaral, 1974]. Thus the study area is the
exposed representative of a geological belt at least 2000 Km
in extent. We propose that this belt marks the limit of the
Amazon craton during the lateMesoproterozoic, deformed in
the studied region by the emplacement of the Paragua block
during the final amalgamation of the Rodinia supercontinent.

2. Regional Geology

[5] The southwestern margin of the present-day Amazon
craton is marked by three different Grenvillian belts: the

Sunsás and Aguapeı́ belts exposed in eastern Bolivia and
along the Bolivia-Brazil border, respectively; and the Nova
Brasilândia belt exposed in southern Rondônia (Figure 1).
The former two belts have been interpreted as delimiting the
Paragua craton, a polycyclic block of at least Paleoprotero-
zoic age and of subcontinental dimensions exposed almost
exclusively in eastern Bolivia [Litherland et al., 1986; Saes,
1999; Geraldes et al., 2001]. The interior of this block is
characterized by the preservation of K-Ar ages and older
structural trends considered to mark the end of cratonization
at circa 1.3 Ga [Litherland et al., 1989]. The Sunsás-
Aguapeı́ belts comprise quartzitic and conglomeratic units
deformed during late Grenvillian (circa 1.0 Ga) times.
Stratigraphic correlations between the Sunsás and Vibosi
groups of Bolivia and the Aguapeı́ Group of Brazil have
been advanced on the basis of lithology [Litherland et al.,
1989; Saes, 1999], with recent SHRIMP dates from the
bottom of the Aguapeı́ sequence exposed in Mato Grosso,
Brazil, suggesting a maximum depositional age of 1231 ±
14 Ma [Santos et al., 2001]. Deformation of the Sunsás belt
is manifested in a network of curvilinear, strike-slip shear
zones, with mylonitic belts observed in an en echelon trend
that generally strikes WNW. Extensive granitic intrusions
accompany the deformation of the Sunsás belt along these
mylonitic belts, followed by many postkinematic granites
[Litherland et al., 1986, 1989]. In contrast, magmatism is
almost entirely absent from the Aguapeı́ belt, which is
marked by greenschist facies mylonites generated during
thrusting and strike-slip deformation along its NNW trend
[Barros et al., 1978] with minor S-type granite magmatism
[Geraldes et al., 2001]. Compared to the Sunsás belt, only
localized deformation affected the Aguapeı́ belt (Figure 2a),
with the external margins of the belt (Serrania Huanchaca in
the west and Rio Branco in the east) still displaying
horizontal or gently folded bedding of preorogenic strata
[Litherland and Power, 1989; Saes, 1999; Fernandes,
1999]. Assuming that the lithostratigraphic correlation
between the units exposed in these localities is valid, the
total width of the Aguapeı́ belt is only �50 km [Saes,
1999].
[6] The Nova Brasilândia Metasedimentary belt lies

�200 km to the NNW from the Aguapeı́ belt in the
Brazilian state of Rondônia. There is no apparent structural
continuity between the E-W trending NBMB and the
northernmost extent of the NNW trending Aguapeı́ belt.
The NBMB was believed to be a preserved Rondôniano-
San Ignacio (1.3–1.5 Ga) feature of the Paragua block, and
was referred to as the Comemoração schists by Litherland et
al. [1986]. Later workers in the Amazon craton incorporated
these conclusions by linking age provinces across the
NBMB, effectively including the Paragua craton as part of
the Amazon craton [e.g., Tassinari and Macambira, 1999;
Tassinari et al., 2000; Santos et al., 2000; Geraldes et al.,
2001]. The deep-water sediments of the NBMB (carbonates
intercalated with fine-grained pelitic and sandy units) are
lithologically distinct from the mature, continental sedi-
ments of the Aguapeı́ and Sunsás Groups. In spite of the
high metamorphic grade of much of the belt (amphibolite to
granulite), locally preserved S0 bedding is still distinguish-

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

2 of 20

TC6004


able in some outcrops, where a repetitive sequence of pelitic
layers interbedded with quartzose layers suggests a turbi-
ditic environment [Rizzotto, 1999]. Metamorphosed mafic
rocks are observed mainly as massive amphibolite units,
which locally preserve mafic cumulate textures. SHRIMP
analysis of detrital zircons from the pelitic rocks in this
sequence yield ages that range from circa 2090 ± 17 Ma to
1215 ± 20 Ma, with the younger group of zircons consid-
ered as a maximum sedimentation age for the sequence
[Rizzotto et al., 1999]. Deformed leucocratic melts injected
along bedding planes from the same locality yield an age of
1122 ± 12 Ma, possibly constraining the timing of meta-

morphism [Santos et al., 2000]. Two episodes of felsic
intrusions characterize much of the NBMB (Figure 2a). The
first set of intrusions, the Rio Branco suite, yields a
conventional U-Pb zircon age of 1098 ± 10 Ma, with
cogenetic gabbros yielding a U-Pb zircon age of 1110 ±
15 Ma [Rizzotto et al., 1999]. Both of these ages are in
agreement with U-Pb SHRIMP ages of 1113 ± 56 Ma
measured from a zircon of the Rio Branco granite [Santos
et al., 2000]. The Rio Branco suite was emplaced syntec-
tonically and is characterized by a gneissic foliation with
locally developed mylonitic fabrics. The second intrusive
suite, known as the Rio Pardo Granite, yields U/Pb ages of

Figure 1. (bottom inset) Map of location of study region (white outline) in South American continent.
Amazon craton is shown in black line, dashed where uncertain. The double black line indicates the
presence of Brasiliano/Panafrican mobile belts and the gray, bold line indicates the location of the
proposed division between the Amazon craton (AC) and the Paragua craton (PC). (top inset) Gray area
outlined in black is area depicted in main image. Bold gray line is the international border between
Bolivia and Brazil. Generalized geometry of ‘‘Grenvillian’’ belt of SW Amazon region shown with
Sunsás belt forming southern boundary of Paragua craton, the limits of which are ill defined to the east
and west. The northern boundary of the Paragua craton is proposed to be the E-W trending Nova
Brasilândia belt, not shown in its full extent here. North of the Nova Brasilândia belt lie the polycyclic
rocks of Amazon craton. Note the disjunctive trends between the Aguapeı́ belt and the Nova Brasilândia
belt. (main) Simplified geological of the Brazilian state of Rondônia and westernmost Mato Grosso.
Older ‘‘Grenvillian’’ deformation at circa 1.2 Ga created the shear zone network in Amazon craton
basement rocks shown in bold gray lines. The boundary between the Amazon craton and the
metasediments of the Nova Brasilândia belt is covered by three sedimentary sequences (Palmeiral
Formation, Primavera Group, and Parecis Formation), which follow the east-west trend of the belt itself.
Bold, black lines in Nova Brasilândia belt indicate major mylonitic zones developed during circa 1.09 Ga
metamorphism.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

3 of 20

TC6004


1005 ± 41 Ma [Rizzotto et al., 1999] and a Rb-Sr isochron
age of 1003 ± 22 Ma (87Sr/86Sr0 = 0.7038, MSWD = 0.3
(C. C. G. Tassinari, Resultados radiométricos pelos métodos
Rb-Sr e K-Ar de rochas do sudoestede Rondônia, internal
report, Co. de Pesquisa de Recursos Minerais, São Paolo),
which are interpreted as a crystallization age. The majority
of these granites display synkinematic to late kinematic
emplacement fabrics, with a small subset intruded at
high crustal levels after the cessation of deformation,
as evidenced by the preservation of miarolitic cavities
(Figure 2b).
[7] The Amazon craton basement lies to the north of

the NBMB and is best exposed in the central portion of
the Brazilian state of Rondônia. The polydeformed base-
ment is marked by the development of a large network of
strike-slip, mylonitic shear zones that display a systematic
sinistral shear sense, postulated to be the result of early
collision of the Amazon craton with southernmost
Laurentia. This network of shear zones was active after
�1.2 Ga under midamphibolite facies conditions and is
observed to overprint both igneous and preexisting meta-
morphic rock fabrics. The late Mesoproterozoic deforma-
tion is the final episode in a complex history of
metamorphism and multiple intrusions that range in age
from circa 1.75 Ga to 1.15 Ga [Bettencourt et al., 1996;
Payolla et al., 2002]. The last magmatic episode that
affected this portion of the Amazon craton was the
emplacement of the tin-bearing Younger Granites of
Rondônia at circa 970 Ma [Bettencourt et al., 1999].

The intrusion of these granites at hypabyssal depths and
their presently undeformed state may mark the final
stages of cratonization in this area.
[8] The actual boundary between the Amazon craton and

the NBMB is covered by three, flat-lying sequences of
continental sediments, attesting to at least three episodes of
reactivation (Figure 1). The Palmeiral Formation is the
oldest stratigraphic unit and comprises pinkish arkosic
sandstones and quartzitic conglomerates with decimeter to
meter scale cross beds [Leal et al., 1978]. Imbricated
structures indicate southward directed paleocurrents for
the deposition of these continental arkoses, the immaturity
of which suggests proximal sources [Bahia, 1997]. The
Palmeiral Formation unconformably overlies the 1198 ±
3 Ma gabbroic sill and subaerial basalts of the Nova Floresta
Formation [Leal et al., 1978; Tohver et al., 2002]. The
contact of the Palmeiral Formation with the metasediments
of the NBMB is an angular nonconformity, suggesting that
deposition postdates the uplift and cooling of the NBMB
[Scandolara and Rizzotto, 1998]. A second episode of
reactivation is marked by deposition of the Primavera
Group into the Pimenta Bueno graben during the Paleozoic
(late Neoproterozoic?). The sediments of this Primavera
Group comprise arenites, glaciogenic diamictites, ferrugi-
nous sandstones, and dolomites that reflect a shallow marine
environment [Leal et al., 1978; Scandolara et al., 1999]. A
final episode of reactivation is marked by late Jurassic
volcanism throughout the region including the emplacement
of kimberlites and alkaline basalts [Montes-Lauar et al.,

Figure 2. Schematic diagram of the tectonostratigraphic framework of the (a) southern boundary of the
Amazon craton together with the adjacent Nova Brasilândia belt and (b) interior of the Paragua craton in
the region of the Aguapeı́ belt in the SE portion of Figure 1.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

4 of 20

TC6004


1994]. Continental sedimentation (Parecis Formation) of
early Cretaceous age followed this volcanism.

3. Structure

[9] Regional mapping and metallogenic prospecting of
the Nova Brasilândia Metasedimentary belt was carried out
systematically by the Brazilian Geological Service (CPRM)
at the 1:100,000 scale in the 7200 km2 region centered
between the towns of Nova Brasilândia d’Oeste and Rolim
de Moura [Scandolara and Rizzotto, 1998; Bahia and Silva,
1998] compiled in Figure 3. Foliations in the NBMB are
defined by the development of both mylonitic shear zones
and metamorphic fabric, generated by mineral growth and
transposition of sedimentary fabrics. Mylonitic foliations
are observed to be both concordant and in a crosscutting
relationship to metamorphic fabrics, suggesting a sequential
history of development discussed below. The metamorphic
fabrics range in intensity from a well-developed schistosity
typical of metapelitic units, to the development of banded

gneisses, which generally characterize calc-silicate units and
higher-grade pelitic rocks. Massive textures characterize
some calc-silicate units as well as the majority of metaba-
sites, with amphibole typically overgrowing a relict cumu-
late texture. Lineations are defined by the presence of a
preferred mineral orientation, generally formed by quartz
ribbons, deformed feldspar augens, or the alignment of
acicular mineral grains [Luft et al., 2000].
[10] A NNE dipping fabric is commonly observed

throughout the northern portion of the mapped region,
characterizing both the attitude of mylonitic shear zones
and regional rock fabrics. Regional strike for these rocks
varies from N60�W to approximately E-W. Fabrics with the
opposite vergence are observed locally in the southern
domain, with more dispersion in regional strike, varying
from N60�E to N60�W. Lineations from these two domains
are characterized by intermediate to high rake angles,
suggestive of a major component of dip-slip motion and
secondary component of strike-slip motion. On a regional-
scale cross section of the western portion of the mapped
region, the NBMB displays a doubly verging structure, with

Figure 3. Map of portion of Nova Brasilândia belt compiled from regional maps (Rio Pardo and Paulo
Saldanha 1:100,000 quadrangles by Bahia and Silva [1998] and Scandolara and Rizzotto [1998].
Schematic cross sections indicate a regional structure with double vergence. Strike-slip mylonite zones
with sinistral offset predominate in the central domain. Age data from U/Pb and 40Ar/39Ar analysis from
this study are shown, indicating regional cooling after circa 1.05 Ga.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

5 of 20

TC6004


the average fabric inclined toward the juncture between
the two domains. Vertical fabrics are observed in the
boundary zone with lineations that define a girdle about
the horizontal plane, suggestive of predominantly strike-slip
motion (Figure 4).
[11] Two deformational phases characterize the structural

evolution of the NBMB, with an initial NE directed com-
pressional event marked by the pervasive development
of the NW trending fabrics [Bahia and Silva, 1998;
Scandolara and Rizzotto, 1998]. Thrust faulting was active

during this phase with tectonic transport to the southwest.
This deformational event seems to have been progressively
followed by the development of E-W trending sinistral,
strike-slip mylonitic zones, since there is not a clear
difference in the metamorphic conditions at which these
events occurred [Luft et al., 2000]. The second deforma-
tional phase was marked by the intrusion of the juvenile
rocks of the Rio Pardo granite suite, which shows a strong
protomylonitic to mylonitic foliation, generally striking
E-W with anastamosing portions striking WNW. The
strike-slip motion is consistent with large-scale sinistral
offsets observed in the Sunsás belt [Litherland et al.,
1986]. Thus it is concluded that both the NBMB and Sunsás
belt were involved in an orogenic zone characterized by
large-scale sinistral strike-slip motion at circa 1.05 Ga
[Litherland et al., 1986, 1989].

4. Metamorphic Petrology

[12] The NBMB is characterized by an increase in
metamorphic grade from north to south. The northern
portion of the sampled region consists largely of impure
quartzites and micaceous, sillimanite-bearing schists, the
more micaceous of which have undergone extensive
weathering. The S0 planes been affected by at least two
generations of folding, as well as the injection of melt
along bedding planes (Figure 5b). Zircons from this
anatectic melt, which is interpreted to be locally derived,
yield a U/Pb age of 1100 ± 8 Ma [Rizzotto et al., 1999].
Most of the mica schist samples are sillimanite-bearing,
but one sample contains the assemblage Ky + St + Sil +
Ms + Bio + Qz + Plag + Ru with the sillimanite appearing
as a late phase growing in biotite (Figure 6a). Calc-silicate
rocks are rare in the northern domain, restricted to small
lenses and tremolite-rich quartzites. Lenses of amphibolite
in the region are of massive aspect, with some evidence of
a relict igneous layering. A typical mineral assemblage
found in the amphibolites is Plag + Hb + Ilm + Mag ±
Cm ± Trm with retrograde chlorite appearing in some
samples (Table 1).
[13] The presence of one- and two-pyroxene granulite

facies rocks occurs only in the southern domain of the
studied region. Calc-silicate rocks are more common than
in the north, appearing as both banded gneisses with
alternating greenish black and pink layers consisting Qz +
Plag + Sph ± Scap + Di interlayered with Hb + Plag + Qz.
Some spectacular examples of these calc-silicate banded
gneiss consist almost entirely of Plag + Gar + Qz + Di +
Sph. Pelitic rocks in the southern domain are characterized
by a gneissic foliation, defined by the alignment of mica
grains. In the highest-grade assemblages, south of the town
of Santa Luzia, mylonitization has imparted a lineation
partially formed by elongate garnet grains with abundant
sillimanite as both inclusions and in the matrix. Fluorine-
and Ti-rich biotite is found as a stable phase in some
high-grade rocks in the southern portion of the NBMB
(Figure 6b), but biotite is generally absent in pyroxene-
bearing rocks of the granulite facies from this southern
domain (Figure 6c).

Figure 4. Lower hemisphere stereonet projections of poles
to foliation and lineations compiled from the Rio Pardo and
Paulo Saldanha quadrangle sheets with contours at 1, 2, 4,
8, and 16% for the (a) northern, (b) central, and (c) southern
domains [Bahia and Silva, 1998; Scandolara and Rizzotto,
1998]. The average foliation pattern in the northern domain
(Figure 4a) is NE dipping, although a small population from
the southern domain displays the opposite vergence,
observed as an isolated population in Figure 4c. Lineations
observed in the northern domain (Figure 4a) are consistent
with dip-slip transport. In the central domain (Figure 4b)
shallow rake lineations are associated with steeply dipping
rock foliations, suggesting strike-slip tectonic transport.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

6 of 20

TC6004


[14] Samples were selected for calculations of pressure
and temperature on the basis of mineral assemblages.
Electron microprobe analysis was undertaken using a
Cameca SX-100 with analyses obtained by comparison
with natural silicate and oxide mineral standards from the
collection of the University of Michigan Electron Micro-
analysis Laboratory. The loci of independent reactions
were calculated using the internally consistent thermody-
namic database of Thermocalc v.3.1. Activity coefficients
were determined for various solid solutions in pyroxene,
plagioclase feldspar, biotite, and hornblende using the A-X
program of Holland and Powell [1998]. The activity of
garnet species was calculated with the asymmetric Qua-
ternary mixing model of Ganguly et al. [1996]. Represen-
tative mineral analyses used for thermobarometric
calculations are contained in Tables 2–6. Sample 326 is
a pelitic schist from the northern portion of the NBMB
with the assemblage Ky + St + Sil + Bi + Ms + Kfs +
Plag + Qz. Kyanite, staurolite, and muscovite appear as
relict grains overprinted by a fine-grained matrix of
sillimanite, K-feldspar, and biotite (Figure 6a). The pres-
ence of muscovite places an additional constraint on

metamorphic conditions, which can be described by three
independent reactions:

1) Al2SiO5 $ Al2SiO5

kyanite $ sillimanite

2) 6Fe4Al18Si7.5O46(OH)2 + 8 KAl3Si3O10(OH)2 + 17
SiO2 $ 8 KFe3Si3AlO10(OH)2 + 62 Al2SiO5 + 12
H2O
Fe-staurolite + muscovite + quartz $ Fe-biotite
+ aluminosilicate + water

3) KAl3Si3O10(OH)2 + SiO2 $ KAlSi3O8 + Al2SiO5 +
H2O
muscovite + quartz $ K-feldspar + aluminosilicate
+ water

The locus of the staurolite breakdown reaction, corrected for
solid solution, is below the terminal reaction, Fe-staurolite +
quartz $ garnet + biotite + sillimanite + H2O, which may
account for the lack of garnet in the rock. Another possible
explanation for the absence of garnet is the highly magnesian

Figure 5. (a) View of typical deformation features in the Aguapeı́ belt registering dynamic
metamorphism under greenschist facies; (b) deformation of the Nova Brasilândia belt accompanied by
partial melting (leucosomatic dike in middle of photograph). In spite of the sillimanite-grade
metamorphism, the turbiditic features of the Nova Brasilândia protolith are well preserved, as evidenced
by the repetitive stratigraphy in the lower portion of the photograph.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

7 of 20

TC6004


bulk composition of the rock, as indicated by an Mg# of
50 for biotite, the modally dominant ferromagnesian phase
(cf. Figure 1 of Thompson [1976]). The coexistence of
kyanite and staurolite indicates early equilibrium under
high P-T conditions. The breakdown of staurolite suggests
an increase in temperature from this early equilibrium
state, given the steep dP/dT that characterizes reaction 2
in the kyanite field,. The breakdown of kyanite is more
likely the result of a decrease in pressure, because further
increases in temperature would have resulted in the growth
of garnet (Figure 7, arrow 1). The presence of a melt
phase in numerous outcrops in the region suggests that the
H2O-saturated solidus for haplogranite may have been
exceeded locally (Figure 7).

[15] Two samples from the southern domain were used to
constrain the P-T conditions of the high-grade portion of the
NBMB. Sample 327.1 is a two-pyroxene garnet granulite
with the assemblage Kfs + Plag + Qz + Gar + Opx + Cpx +
Ilm + Ap + Zc. Hornblende is observed as overgrowths on
pyroxene, and so is not considered to be part of the peak
assemblage. Garnets are subhedral to euhedral, with com-
positional profiles characterized by an increase (�4 mol%)
in grossular content toward the rims (�50 mm thick), but
little change in Mg#. However, Mg# does vary from one
garnet grain to another, with the highest values preserved in
grains distant from other ferromagnesian phases, possibly
reflecting a lack of resetting upon cooling. Inclusions in
garnet include clinopyroxene, orthopyroxene, plagioclase

Figure 6. (a) Typical pelitic assemblage (sample 326) from the low-grade, northern portion of the Nova
Brasilândia belt. Observe the relict staurolite and kyanite overprinted by a fine-grained mesh of
sillimanite and biotite. (b) High-grade pelitic assemblage (sample 134) from the southern portion of the
belt with mylonitic foliation. Gray inclusions in elongate garnet are spinel, with both sillimanite and
ilmenite found as inclusions and in the matrix. Rutile is found only as a matrix phase. (c) Two-pyroxene
garnet granulite (sample 327.1) with clinopyroxene inclusion in garnet. (d) Example of retrograde Mg-Fe
halo surrounding clinopyroxene inclusion in garnet. Black lines are grossular contours (20–24 mol%),
while Mg# contours are drawn at 0.5 intervals (gray lines). Note the difference in Mg# for the garnet grain
close to matrix pyroxene and garnet profile in Figure 8, taken from a garnet removed from other
ferromagnesian phases. Scale bar is 1 mm for all samples, except for Figure 6d, where scale bar is 100 mm.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

8 of 20

TC6004


Table 1. Representative Mineral Assemblagesa

Sample Description Gar Bi Als Ms Qz Plag Kfs Hb Amp St Chl Ru Ilm Sph Opx Cpx Other

ron127 mica schist . . . x . . . x x x . . . . . . . . . . . . . . . . . . x . . . . . . . . . . . .
ron 128 amphibolite . . . . . . . . . . . . . . . x . . . x Cm . . . . . . . . . x . . . . . . . . . . . .
ron 130 amphibolite . . . x . . . . . . . . . x . . . x . . . . . . r . . . x . . . . . . . . . . . .
ron 131 metabasite . . . x . . . . . . . . . x . . . x Trm . . . x . . . x . . . . . . . . . mt
ron 132 calc-silicate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . scp
ron 133 metabasite . . . . . . . . . . . . x x . . . x . . . . . . . . . . . . . . . . . . x x . . .
ron 134 mylonitic gneiss x x Sil x x x x . . . . . . . . . . . . x x . . . . . . . . . Mon, Zc, Gr
ron 135b amphibolite . . . . . . . . . . . . x . . . x . . . . . . . . . . . . x x . . . . . . . . .
ron 135a calc-silicate . . . . . . . . . . . . x x . . . x . . . . . . . . . . . . . . . x . . . . . . Scap
ron 137 granitoid . . . x . . . . . . x x x . . . . . . . . . . . . . . . x . . . . . . . . . Zc
ron 138 mica schist . . . . . . . . . x x x x . . . . . . . . . . . . . . . . . . . . . . . . . . . Mon
ron 139 mica schist . . . x . . . . . . x x . . . . . . . . . . . . . . . x x . . . . . . . . . . . .
Ron 316 calc-silicate . . . . . . . . . . . . x x . . . . . . . . . . . . . . . . . . . . . x . . . x Scp
ron 317 granitoid . . . x . . . x x x x . . . . . . . . . . . . . . . x . . . . . . . . . All
ron 319 calc-silicate . . . . . . . . . . . . x x . . . . . . Trm . . . . . . . . . . . . x . . . x . . .
ron 320 amphibolite . . . x . . . x x x . . . . . . . . . . . . . . . x x . . . . . . . . . . . .
ron 322.2 calc-silicate . . . . . . . . . . . . x x . . . x . . . . . . . . . . . . . . . . . . . . . x . . .
ron 324d mica schist . . . x Sil x x x . . . . . . . . . x . . . x . . . . . . . . . . . . Mon,Zc
Ron 325 mica schist . . . x Sil x x x . . . . . . . . . . . . r . . . x . . . . . . . . . Mon,Zc
ron 326 mica schist . . . x Ky, Sil x x x . . . . . . . . . x . . . . . . . . . . . . . . . . . . Mon
ron 327.1 garnet granulite x . . . . . . . . . x x x . . . . . . . . . . . . . . . x . . . x x Zc,FeS
ron 328 amphibolite . . . . . . . . . . . . x x . . . x . . . . . . r . . . x x . . . . . . . . .
Ron 329 granulite . . . . . . . . . . . . x x x . . . . . . . . . . . . . . . . . . . . . x x Zc
Ron 401b mica schist . . . x . . . x x x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zc
Ron 404c mica schist . . . x . . . x x x . . . . . . . . . . . . . . . . . . x . . . . . . . . . Mon
Ron 404d mica schist . . . . . . . . . . . . x X . . . . . . Grn . . . . . . . . . . . . x . . . . . . Zc
Ron 404f mica schist . . . x Sil x x X . . . . . . . . . . . . . . . x x x . . . . . . Zc

aAbbreviations are as follows: Gar, garnet; Bi, biotite; Als, aluminosilicate; Ms, muscovite; Qz, quartz; Plag, plagioclase; Kfs, K-feldspar; Hb,
hornblende; Amp, other amphiboles; St, staurolite; Chl, chlorite; Ru, rutile; Ilm, ilmenite; Sph, titanite; Opx, orthopyroxene; Cpx, clinopyroxene; Sil,
sillimanite; Ky, kyanite; Mic, microcline; Cm, cummingtonite; Trm, tremolite; Grn, grunerite; Mag, magnetite; Scp, scapolite; Gr, graphite; All, allanite;
Mon, monazite; Zc, zircon; r, retrograde.

Table 2. Pyroxene Analysesa

Sample
327 Opx
Matrixb

327 Cpx
Matrixb

327 Opx
Inclusionb

327 Cpx
Inclusionb

SiO2 48.74 50.74 48.45 50.48
TiO2 0.02 0.01 0.01 0.01
Al2O3 0.16 1.02 0.53 1.19
Cr2O3 0.06 0.01 0.00 0.00
FeO 39.67 17.83 40.51 17.31
MnO 0.46 0.21 0.51 0.18
MgO 10.22 8.43 8.98 8.20
CaO 0.86 21.14 0.88 20.92
Na2O 0.01 0.35 0.02 0.35
Sum 100.17 99.72 99.86 99.55
Si 1.98 1.97 1.98 1.97
AlIV 0.02 0.03 0.02 0.03
AlVI 0.00 0.02 0.01 0.02
Ti 0.00 0.00 0.00 0.00
Cr 0.00 0.00 0.00 0.00
Fe3+ 0.04 0.02 0.01 0.03
Fe2+ 1.30 0.56 1.38 0.56
Mn 0.02 0.01 0.02 0.01
Mg 0.62 0.49 0.55 0.48
Ca 0.04 0.88 0.04 0.87
Na 0.00 0.02 0.00 0.02
A(en) 0.081 . . . 0.102 . . .
A(fs) 0.48 . . . 0.43 . . .
A(di) . . . 0.41 . . . 0.4
A(hed) . . . 0.5 . . . 0.51

aNormalized to four cations.
bReintegrated from 30–50 spot analyses.

Table 3. Garnet Analysesa

327.1 Core 327.1 Rim 134 Core 134 Rim

SiO2 37.66 37.79 37.37 38.54
TiO2 0.01 0.01 0.19 0.01
Al2O3 21.07 21.22 22.47 22.40
Cr2O3 0.00 0.00 0.17 0.36
FeO 32.02 30.57 29.68 30.48
MnO 1.36 1.22 0.38 0.26
MgO 2.58 2.42 8.01 8.09
CaO 6.80 7.91 1.36 1.20
Na2O 0.02 0.01 0.03 0.06
Sum 101.49 101.15 100.41 101.40
Si 2.97 2.98 2.90 2.96
Ti 0.00 0.00 0.01 0.00
Al 1.96 1.97 2.05 2.02
Cr 0.00 0.00 0.01 0.00
Fe3+ 0.04 0.02 0.04 0.02
Fe2+ 2.11 2.02 1.93 1.96
Mn 0.09 0.08 0.02 0.02
Mg 0.30 0.28 0.92 0.92
Ca 0.57 0.67 0.11 0.10
Alm 67.8 65.6 63.8 65.1
Gros 18.4 21.7 3.7 3.3
Pyr 9.7 9.3 30.0 30.0
Spes 2.9 2.6 0.8 0.5
Uvar 0.0 0.0 0.3 0.0
mg# 12.5 12.4 32.4 32.1
a(gr) 0.0113 0.0200 0.0003 0.0001
a(pyr) 0.0033 0.0033 0.0421 0.0325
a(alm) 0.2962 0.2536 0.2799 0.3316

aNormalized to eight cations.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

9 of 20

TC6004


(An40-An42), and quartz. Large plagioclase grains in the
matrix are zoned as well, with core compositions of An45
decreasing to An35 toward the rims (�75 mm thick),
suggesting that plagioclase inclusions in garnet record
conditions similar to those preserved in matrix plagioclase
cores (Figure 8). Clinopyroxene and orthopyroxene are
present as euhedral grains that lie within the foliation plane.
Exsolved textures in pyroxene reflect unmixing during slow
cooling, with original compositions recovered through
standard reintegration techniques [Bohlen and Essene,
1977]. Clinopyroxene inclusions in garnet are surrounded
by Mg-Fe diffusion profiles, probably reflecting ongoing
Mg-Fe exchange during retrograde cooling. In order to
recover peak temperatures from the inclusions, the area of
the haloes was determined by image analysis and excess
Fe from this area was reintegrated with the composition
of inclusions (Figure 6d). The reintegration yields temper-
atures that were higher by �50�C than temperatures
obtained without reintegration. Thermobarometric calcula-
tions (Figure 7) for the garnet core and inclusion assemblage
gives 810�C at pressures of 7.5 MPa (SdT 150, SdP 1.8,
Sdfit 0.04, correlation 0.90), derived from garnet-pyroxene
exchange reactions as well as garnet-pyroxene-plagioclase
mass transfer reactions (e.g., GADS [Moecher et al., 1988]).
[16] In using the core-matrix and rim-contact assemblages

to calculate P-T conditions, it is uncertain whether zoning in
garnet and plagioclase is due to prograde changes in pressure
or to retrograde reactions such as the late appearance of
hornblende or K-Na exchange between K-feldspar and
plagioclase. Given the possibility of retrograde, mass-
transfer and exchange reactions, the pressure increase of
�3 kbar indicated by growth zoning in garnet and plagio-

Table 4. Staurolite Analysesa

326 Mean 326 Core 326 Rim

SiO2 26.92 26.47 26.65
Al2O3 55.28 55.95 55.37
TiO2 0.47 0.51 0.48
FeO 14.37 14.77 14.35
MgO 1.58 1.60 1.63
MnO 0.54 0.46 0.59
ZnO 1.93 1.96 2.10
Na2O 0.06 0.02 0.06
Cl 0.01 0.01 0.00
F 0.03 0.03 0.00
Sum 101.24 101.77 101.22
Si 7.66 7.52 7.59
AlIV 0.34 0.48 0.41
AlVI 18.18 18.23 18.17
Ti 0.10 0.11 0.10
Fe3+ 0.11 0.11 0.11
Fe2+ 3.30 3.37 3.29
Mg 0.67 0.68 0.69
Zn 0.40 0.41 0.44
Ca 0.00 0.00 0.00
Na 0.03 0.01 0.03
Cl 0.01 0.01 0.00
F 0.06 0.05 0.00
a(fst) 0.43 0.43 0.43
A(mst) 0.013 0.015 0.013

aAssuming 3.5% Fe3+, normalized to 48 O atoms.

Table 5. Sheet Silicate Analysesa

127 Matrix 134 Contact
134

Matrix
326

Biotite
326

Muscovite

SiO2 36.64 36.16 35.80 36.64 48.48
TiO2 2.43 1.87 4.21 1.69 0.12
Al2O3 18.72 17.55 16.08 20.40 37.57
Cr2O3 0.00 0.19 0.16 0.00 0.00
FeO 18.93 8.90 14.30 19.50 1.18
MgO 11.47 18.72 13.54 10.25 0.79
MnO 0.22 0.14 0.00 0.18 0.03
BaO 0.00 0.19 0.00 0.00 0.00
CaO 0.03 0.00 0.01 0.02 0.00
Na2O 0.13 0.13 0.03 0.14 0.27
K2O 9.92 10.38 10.46 9.73 9.68
F 0.38 2.29 0.63 0.33 0.11
Cl 0.11 0.14 0.21 0.10 0.02
Sum 98.30 95.66 94.28 98.38 98.25
Si 5.49 5.21 5.56 5.48 6.18
AlIV 2.51 2.78 2.44 2.52 1.82
AlVI 0.79 0.19 0.51 1.08 3.81
Ti 0.27 0.20 0.49 0.19 0.01
Cr 0.00 0.02 0.02 0.00 0.00
Fe3+b 0.19 0.08 0.15 0.19 0.11
Fe 2.16 0.98 1.69 2.23 0.00
Mg 2.56 4.02 3.14 2.29 0.15
Mn 0.03 0.02 0.00 0.02 0.00
Ba 0.00 0.01 0.00 0.00 0.00
Ca 0.00 0.00 0.00 0.00 0.00
Na 0.04 0.04 0.01 0.04 0.07
K 1.90 1.91 2.07 1.86 1.57
F 0.18 1.04 0.31 0.16 0.04
Cl 0.03 0.03 0.06 0.03 0.00
Mg* 54.1 80.2 64.9 50.3 98.0

aNormalized to 22 oxygen atoms.
bEstimated from Guidotti and Dyar [1991].

Table 6. Oxide Analysesa

134 Ilmenite 134 Spinel

FeO 43.95 22.49
TiO2 51.56 0.03
Al2O3 0.02 62.66
MnO 1.15 9.50
MgO 0.03 0.02
ZnO 0.05 5.28
Cr2O3 0.05 0.22
SiO2 0.02 0.03
*V2O3 0.55 0.34
NiO 0.00 0.00
Sum 97.38 100.57
Fe3+ 0.00 0.00
Fe2+ 0.95 0.51
Ti 1.01 0.00
Al 0.00 1.99
Mn 0.03 0.00
Mg 0.00 0.38
Ca 0.00 0.00
Cr 0.00 0.00
Si 0.00 0.00
Vb 0.01 0.01
Zn 0.00 0.11

aNormalized to two cations.
bV corrected for Ti interference in ilmenite.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

10 of 20

TC6004


clase calculated using the garnet-pyroxene-plagioclase-
quartz geobarometers of Moecher et al. [1988], probably
represents a maximum estimate. Average P-T conditions
derived from Thermocalc 2.1 using the same reactions as
before (Figure 7) indicate a slight decrease in temperature
and increase in pressure from conditions recorded by the
core assemblage (arrow 2, Figure 7), 780�C at 8 MPa (SdT
110, SdP 1.3, Sdfit 0.7, correlation 0.83).
[17] Sample Et-134 is a sillimanite-garnet gneiss with the

assemblage Gar + Sil + spinel + Plag + Qtz + Bi + Ru + Ilm.
Sillimanite, spinel, quartz, ilmenite, and plagioclase (An28-
An31) form large inclusions within the garnet and are also
found as matrix phases. Rutile is observed only in the
matrix, typically in close association with ilmenite. Plagio-
clase in the mylonitic matrix of the sample ranges in
composition from An20 to An25. Garnet profiles are homo-
geneous with a slight decrease (1 mol%) in grossular
content and no systematic changes in Mg#. Pressures and
temperatures calculated for the coexisting core-inclusion
assemblage are 770�C at 8 MPa (SdT 140, SdP 1.3 MPa,
Ssdfit 1.2, correlation 0.81), based on GASP and garnet-
biotite reactions. The rim-matrix assemblage for sample 134
(Gar + Plag + Qz + Sil + Bi + Ru + Ilm) yielded average
temperatures of 630�C at 6 MPa (SdT 130, SdP 1.5, Sdfit

1.9, correlation 0.62) based on GRIPS, GRAIL, and GASP
geobarometers and garnet-biotite thermometry. The P-T
differences between core and rim assemblages indicate
cooling and decreasing pressures (arrow 3, Figure 7). There
is no textural evidence for late kyanite in the retrograde
assemblage, as predicted by the calculated P-T conditions.
[18] In summary, the metamorphic conditions for the

northern and southern domains show distinct variations in
grade and in evolving P-T conditions. A prograde sequence
is preserved in the northern domain, where metamorphic
conditions never reached the granulite facies, with the
breakdown of kyanite consistent with decreasing pressure.
The clockwise loop implied by this sequence is typical of
terranes that have undergone crustal thickening due to
imbrication [England and Thompson, 1984]. In contrast,
metamorphic conditions in the granulitic southern domain
register increasing pressures, coupled with a slight decrease
in temperature, followed by exhumation (counterclock-
wise loop). The difference in evolving P-T conditions
observed along a N-S transect reflect separate processes
by which orogenesis can result in crustal thickening;
imbrication through deep seated thrust faulting and the
development of a volcanic overburden caused by melt
extraction from deeper, granulitic source rocks, a process

Figure 7. Compiled P-T information for the northern and southern portions of the Nova Brasilândia
belt. Temperature limits on the prograde sequence for the northern domain are indicated by two staurolite-
out reactions (solid black lines crossing AlSi2O5 phase equilibria with boldface text) [Dutrow and
Holdaway, 1989]. Dotted lines indicate representative reactions used for sample 327 (squares, shaded
error ellipses with dotted lines), and dashed lines indicate reactions used for sample 134 (circles, open
error ellipses with dashed lines). P-T conditions for solid symbols are from core-inclusion assemblages
and open symbols are from rim-contact assemblages. Top inset shows separate P-T loops for both
portions of the belt, with uniform cooling indicated by single path with geochronological constraints for
cooling through hornblende’s blocking temperature.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

11 of 20

TC6004


which transfers heat to rocks stratigraphically higher in
the crustal pile, driving prograde reactions [e.g., Valley,
1992]. That both processes were at work in the trans-
pressional NBMB suggests a tectonic collision with some
degree of obliquity.

5. Geochronology

[19] Samples were collected from throughout the
NBMB for both 40Ar/39Ar and U/Pb geochronology.
Argon analysis was performed at the University of

Michigan (see Tohver et al., 2002 for methods) and
U/Pb analysis was undertaken at the Zentrallabor für
Geochronologie at the Universität Münster, in Münster,
Germany. Rocks were crushed and sieved for the 100–
500 mm size fraction, which was washed with deionized
water, dried, and passed through heavy liquids to remove
lighter fractions. A Frantz magnetic separator was used
to divide diamagnetic and paramagnetic fractions. Individual
titanite and monazite grains were hand-picked under
a binocular microscope, avoiding grains with visible
inclusions.

Figure 8. Mineral zoning patterns in (top) garnet and (bottom) plagioclase two-pyroxene garnet
granulite (sample 327.1), suggesting garnet growth at the expense of the anorthitic component of
plagioclase under increasing pressures.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

12 of 20

TC6004


[20] Mineral fractions were washed, weighed, and spiked
with a mixed 233U-205Pb tracer before being placed in 3 mL
Teflonk vials for digestion (HF and HNO3 for titanite,
HNO3 and HCl for monazite). The sample vial were placed
in Teflonk-lined, Krogh style bombs at 200�C for 24–
72 hours, after which the dissolved samples were dried.
Both U and Pb were separated in 100 mL Teflonk columns
with Dowex 1 � 8 resin beads using HCl-HNO3 and
HCl-HBr anion exchange chemistry, respectively. Dried U
and Pb fractions were redissolved in concentrated HCl and
loaded onto separate Re filaments using the phosphoric
acid-silica gel technique. Fractionation during thermal
ionization was corrected by measurement of Pb standard
NBS-982 and the U standard U-500. Both U and Pb were
analyzed for isotopic composition using a thermal ionization
VG Sector 54 multicollector mass spectrometer. Most
masses were measured in Faraday cups with 204Pb measured
on the Daly detectors for samples with low amounts of
common Pb.
[21] Monazite and titanite grains from six samples

throughout the metasedimentary belt were selected for
U/Pb analysis (Table 7). Results from these analyses are
plotted on U-Pb concordia diagrams in Figure 9. The inter-
cept for discordant aliquots was determined from a line
forced through the origin. Most monazite samples are char-
acterized by radiogenic Pb in excess of 98% of total Pb, but
titanite shows significant incorporation of common Pb, with
206Pb/204Pb ratios as low as 113 (85% radiogenic Pb).
Common Pb correction for all samples was made using
Stacey and Kramer’s two-stage Pb evolution curve (1975)
with values at 1.0 Ga of 17.06 for 206Pb/204Pb and 15.51 for
207Pb/204Pb (Table 7). The resultant titanite ages are similar
for all samples, regardless of the amount of common Pb
initially incorporated, suggesting that isotopic values of
initial Pb values were reasonably uniform for the region.
Two monazite ages are concordant (sample 134, U/Pb age
1095.8 ± 5.3 Ma; and sample 319, U/Pb age 993 +11/�12 Ma)
with one sample plotting slightly above the U-Pb concordia
(sample 326, 207Pb/206Pb age 1082.4 +6.2/�6.4 Ma). The
�100 Ma variation in monazite ages is difficult to explain,
if one assumes that all of these ages represent simple
cooling. One possibility for the youngest age of �993 Ma

is that monazite grew below its closure temperature through
dynamic recrystallization. This scenario is supported by the
observation that most titanite, which has a lower closure
temperature than monazite, yields ages in the range of
1020–1084 Ma, which is older than the youngest monazite.
Most titanite grains are 93–97% concordant with the upper
limit on the age range of titanite defined by sample GR-11,
which has a relatively low 206Pb/204Pb ratio of 113. The
dependence of this sample on the common Pb correction
renders the 7.5% discordant age less reliable.
[22] Continued cooling from the peak of regional meta-

morphism is recorded by hornblende ages that cluster in the
�970 Ma range and biotite ages in the �940 Ma range.
Argon isotope data are available as auxiliary material1.
Spectra from step-heating of minerals for 40Ar/39Ar analysis
are shown in Figure 10. As can be seen from the degassing
spectra, after the initial, low temperature steps, most sam-
ples exhibit undisturbed plateaus characterized by a minimal
compositional variation, suggestive of a simple cooling
history with little indication of mineralogical change during
retrograde metamorphism. The geographic distribution of
all age data is shown in Figure 3. The regional distribution
of ages shows no correlation with metamorphic grade,
suggesting that the studied portion of the Nova Brasilândia
was exhumed uniformly. Cooling from the high-grade event
was on the order of 2�–2.5�C/Ma, rates that are typical of
high-grade terranes of the North American Grenville Prov-
ince [Cosca et al., 1991; Mezger et al., 1992]. However, the
large difference between the titanite ages and hornblende
ages, as well as the reset 996 Ma monazite age suggests that
there may have been a phase of deformation at 1.0 Ga.

6. Discussion

[23] The Nova Brasilândia metasedimentary belt records
the effects of high-grade metamorphism at circa 1.09 Ga.
The sequence of metamorphism preserved in the northern
domain suggests a clockwise P-T loop, the product of
crustal thickening through imbrication [England and

Table 7. U/Pb Data for the Nova Brasilândia Belta

Sample U, ppm Pb, ppm 206Pb/204Pb 208Pb/206Pb 207Pb/206Pb 207Pb/235U 206Pb/238U Rho

Age Data, Ma

206Pb/238U 207Pb/235U 207Pb/206Pb

16 t. 14.65 6.16159 164.5 1.367 0.07323 ± 0.0005 1.705 ± 0.015 0.1688 ± 0.0007 0.658 1006 1011 1020
ron135 t. 0.4355 0.1096 252.8 0.5048 0.07469 ± 0.0005 1.769 ± 0.039 0.1718 ± 0.0036 0.954 1022 1034 1060
319 mon 79.59 22.16 239.0 0.5638 0.07225 ± 0.0004 1.674 ± 0.025 0.1680 ± 0.0022 0.935 1001 999 993
134 mon 4600 3592 9736 3.801 0.07599 ± 0.0001 1.935 ± 0.008 0.1847 ± 0.0007 0.936 1093 1093 1095
134 mon 176.9 138.5 10562 3.801 0.07607 ± 0.0001 1.942 ± 0.007 0.1851 ± 0.0006 0.920 1095 1096 1097
GIL t . . . 31.88 839.0 0.6011 0.07463 ± 0.0002 . . . . . . . . . . . . . . . 1059
326 mon 402.4 170.9 877.0 1.518 0.07552 ± 0.0002 1.918 ± 0.008 0.1842 ± 0.0006 0.861 1090 1088 1082
Gr13 t 9.891 3.005 700.0 0.5712 0.07503 ± 0.0002 2.096 ± 0.022 0.2026 ± 0.0020 0.977 1189 1148 1069
Gr11 t 146.2 51.95 112.9 0.6661 0.07560 ± 0.0007 1.756 ± 0.021 0.1684 ± 0.0008 0.626 1004 1029 1085

aAbbreviations are as follows: mon, monazite; t, titanite. Multigrain aliquots were handpicked from heavy liquid separates using a binocular microscope.
Monazite analyses comprise 3–5 of �50–300 mm grains and titanite analyses 0.5–5 mg aliquots of �30–500 mm grains. All errors are calculated at the 2s
level. Mass fractionation was determined to be 1.001 per atomic mass unit and analytical blanks ranged from 30–120 picograms Pb.

1Auxiliary material is available at ftp://ftp.agu.org/apend/tc/
2003TC001563.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

13 of 20

TC6004


Thompson, 1984]. Assuming an average crustal thickness of
30–35 km, the exhumation of the NBMB rocks from
midcrustal depths (20–30 km) implies that metamorphism
was the product of collision-related deformation. The
deformation is registered in the development of deep-seated

thrusting and accompanying strike-slip motion, indicating a
transpressive tectonic environment. The absence of older
basement rocks that have been reworked during deforma-
tion means that the Nova Brasilândia belt records a single
orogenic cycle during late Mesoproterozoic times. This

Figure 9. U-Pb concordia diagrams for monazites and titanites from Nova Brasilândia.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

14 of 20

TC6004


Figure 10. The 40Ar/39Ar age spectra for minerals from the Nova Brasilândia metasedimentary belt
showing homogeneous mineral compositions and insignificant variation in ages measured from replicate
samples.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

15 of 20

TC6004


monocyclic history contrasts strongly with the older, poly-
cyclic Amazon basement to the north, where basement
rocks were being actively deformed in a strike-slip envi-
ronment as recently as 1.13 Ga. The lack of evidence in the
NBMB for the older deformation episode that affected the
cratonic basement to the north has a simple explanation:
the sedimentary protolith of the NBMB was not in place
until after the last tectonometamorphic episode that affected
the Amazon basement rocks at 1.19–1.13 Ga. This scenario
is in agreement with the maximum limits on the NBMB
depositional age, constrained by 1211 ± 18 Ma age of the
youngest detrital zircon present in the metamorphosed
greywackes [Santos et al., 2000].
[24] The location of the NBMB on the boundary zone

marking the limit of the occurrence of the granitoid base-
ment rocks of the Amazon craton becomes significant in
light of aeromagnetic data from this region. Aeromagnetic
data from the SW Amazon craton were collected in two
stages, 1975–1979 and 1988–1995. Total field magnetic
surveys were conducted by airborne proton precession
magnetometer surveys from altitudes that ranged from
150 m for the first phase of data collection to 1020 m,
for the second phase. Field measurement were taken at
intervals of 1–2 s, translating to a ground spacing of
measurements of 60–120 m. Transects in a N-S direction
were spaced over 1 km intervals for the first phase, with up
to 20 km intervals for the second phase. The first phase of
collection was undertaken by the Companhia de Pesquisa
de Recursos Minerais (CPRM) under contract to the
Departamento Nacional de Produção Mineral (DNPM)
and the second phase was undertaken by Petrobras. These
data sets have been made available by the CPRM at http://
www.cprm.gov.br. A compilation of the results of seven
projects (Pacaás Novos (1978), Serra dos Parecis, Cabeceira
do Rio Guaporé, Rio do Sangue, Reserva Indı́gena Juruena,
Bacia dos Parecis, and Bacia dos Parecis (sub Bacia Alto
Xingu)/Bloco 1) is shown in Figure 11.
[25] The subsurface geology of the exposed Nova Brasi-

lândia belt is characterized by two strong magnetic gradients
that outline the northern and southern boundaries of the belt.
The magnitude of the anomalies is consistent with the
presence of small, mafic bodies at depth, such as the
amphibolitized sills and dikes that are exposed throughout
the belt. The general E-W trend of the anomalies is
concordant with the trend of regional geology. This E-W
trend is also observed to mark the southern boundary of the
polycyclic Amazon basement (central portion of Figure 11),
which is characterized by a strong positive anomaly with a
wavelength of �30 km. The amplitude of the associated
negative anomaly to the S is attenuated where covered by
the sediments of the Pimenta Bueno graben, but becomes
more pronounced to the E. This strong anomaly probably is
probably associated with the presence of mafic rocks [e.g.,
amphibolitic bodies of Figure 2] along the 1000 km length
of the E-W trending, first order feature (i.e., the Ciriququi
mafic arc of Litherland et al. [1986]). The anomaly is
continuous for 600 km beneath the continental sediments
of the Serra dos Parecis Formation (Kt) until the nearly
orthogonal intersection with the Brasiliano-aged Paraguai

belt, recognized to form the boundary of the Amazon craton
by late Neoproterozoic times [Alvarenga and Trompette,
1993; Alvarenga et al., 2000]. It is noted that the E-W
magnetic lineament that defines the length of the NBMB
has served as the locus for at least three episodes of
sedimentation as described in a previous section. In addi-
tion, the regional trend of kimberlite occurrences overlaps
with the magnetic lineament, suggestive of a zone of crustal
weakness reactivated during sedimentation and later
exploited during kimberlite emplacement [Wilding et al.,
1991; Svisero, 1995].
[26] The concordance of magnetic anomaly and surface

geology of the adjacent NBMB suggest that the two are
correlated and the belt extends �1000 km farther to the east
than its present exposure. Age correlations from borehole
samples �750 km to the west suggest an overall E-W length
of �2000 km. In contrast, the Aguapeı́ belt (NNW trend) is
a minor map feature with no clear linkage to the Nova
Brasilândia belt. The disjunction between these two meta-
sedimentary belts suggests that they are not connected by
contiguous lithologies and were not tectonically linked, an
interpretation that is corroborated by geochronological data
from the two belts. Whereas the 40Ar/39Ar ages for biotite
from the Nova Brasilândia belt reported in this study are
similar to K/Ar ages of white mica from the Aguapeı́ belt
reported by Geraldes et al. [1997]; the former reflect cool-
ing from a high-grade deformation event at 1.09 Ga, the
latter reflect the absolute timing of deformation under lower
greenschist facies conditions. In the latest Mesoproterozoic,
the NBMB was marked by the deposition of deep-water
sediments (<1215 Ma) followed by high-grade deformation
and metamorphism at �1.09 Ga. The deformation of the
cratonic basement to the north (1.18–1.15 Ga) and concur-
rent sedimentation in the NBMB suggests a genetic link. We
propose that this boundary represents the margin for the
Amazon craton prior to the late Mesoproterozoic accretion
of the Paragua craton to the Amazon craton. How far apart
the Amazon and Paragua cratons were prior to this collision
and whether they share a geological history prior to the late
Mesoproterozoic is unknown.
[27] Regional stratigraphic correlations are an important

test of the hypothesis that the NBMB is a suturing colli-
sional belt. Thus the question of whether the NBMB
sedimentary protolith sequence is related to the Aguapeı́
Group is important to resolving the paleogeographic posi-
tion of the Amazon craton relative to the Paragua craton.
Unrelated stratigraphies would imply no geographic prox-
imity between the respective depocenters, a scenario per-
mitted by the suture hypothesis. On the other hand,
correlating the sequences would signify that the Amazon
and Paragua cratons were geologically linked well before
the end of Mesoproterozoic times. One current model for
the development of the Nova Brasilândia belt interprets the
sedimentary protolith of the belt as having been deposited in
a rift between the Amazon craton and the Paragua craton
[Rizzotto, 2001]. The rift is assumed to extend 400 km to the
SE, encompassing the Aguapeı́ sediments and implies a
tectonic connection between the Amazon and Paragua
cratons. The Aguapeı́ Group itself has been assigned an

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

16 of 20

TC6004


aulacogenic origin, on the basis of stratigraphic columns
that are symmetrical with respect to the central axis of the
basin, as well as paleocurrent analysis that demonstrates
initial transport toward the central axis [Saes, 1999]. U-Pb
ages for detrital zircons for both the Nova Brasilândia and
Aguapeı́ sequences do permit the proposed correlation, with
maximum ages of deposition constrained to 1215 ± 20 Ma
and 1231 ± 14 Ma, respectively [Santos et al., 2000].
However, these temporal limits do not serve as prima facie
proof of a geographic connection between the two belts.
Given the clear differences in tectonic setting between the
Nova Brasilândia sediments (calcareous turbidites and grey-
wackes) and the Aguapeı́ sediments (quartzitic conglomer-
ates and arenites), we are convinced that the correlation is
unwarranted.
[28] The effects of collisional tectonics registered in the

NBMB suggest a different scenario for the deposition of the
Nova Brasilândia protolith. In contrast with the symmetrical
rifting and passive tectonic setting implied above, an alter-
native model is that the sediments register basin formation
in response to an active tectonic environment, either as a

foreland basin or as a transtensional basin in a strike-slip
environment. In the former case, an orogen-parallel basin
responding to crustal loading in advance of a foreland thrust
belt would be filled with sediments shed from tectonically
uplifted regions. In this model, the burial and metamorphism
of the Nova Brasilândia flysch deposits marks the closing of
the foreland basins in the final phases of collision. We prefer
the second model for the deposition of the Nova Brasilândia
protolith sequence, drawing support from observations of
sinistral strike-slip deformation that affected the Amazon
craton at 1.18–1.15 Ga as well as the transpressional ele-
ments (thrust sheets coupled with sinistral strike-slip faults)
in the younger deformational history of the NBMB itself. In
this model, strike-slip motion past an irregular cratonic
margin marked by promontories and reentrants would result
in alternating episodes of transpression and transtension
(Figure 12). Thus the opening of the Nova Brasilândia
seaway can be interpreted as a transtensional phase marking
the passage of the Amazon craton past the southeastern
promontory of Laurentia resulting in the development of a
large releasing bend (Figure 12). The subsequent collision of

Figure 11. Aeromagnetic anomaly data for the SW Amazon craton. The Nova Brasilândia trend is
marked by a clear 1000 km long anomaly that is covered in the eastern extent by the Paleozoic sediments
of the Pimenta Bueno graben (labeled Pz) and Cretaceous sediments of the Parecis Formation (labeled
Kt). Locations of known kimberlite pipes are designated with white stars [Companhia de Pesquisa de
Recursos Minerais, 1983a, 1983b, 1983c; ENCAL, 1989; GEOMAG, 1995, 1996; PROSPEC, 1978].
(ENCAL report is archived in SEDOC/PETROBRAS headquarters in Rio de Janeiro under registration
number 103–07407; GEOMAG [1995] report is archived in the SEDOC/GEREX/E&P headquarters in
Rio de Janeiro under registration number 101–09824; GEOMAG [1996] report is archived in the
SEDOC/GEREX/E&P headquarters in Rio de Janeiro under registration number 101–09974.) See color
version of this figure at back of this issue.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

17 of 20

TC6004


the Amazon craton with the greater Paragua craton (i.e.,
basement rocks on both sides of the Aguapeı́ aulacogen)
resulted in the crustal thickening and metamorphism ob-
served at circa 1.09 Ga. In contrast, if the Aguapeı́ belt is a
closed aulacogen, the Paragua craton extend much farther to
the east, terminating in the Brasiliano-aged Paraguai belt.
[29] Given the proposed tectonic links between Amazo-

nia and Laurentia, the tectonic history of the ‘‘Grenvillian’’
margin of Amazonia has implications for Rodinia paleoge-
ography. The strike slip mylonitic shear zones preserved on
the Amazon side and the recent discovery of Grenvillian
terranes with Transamazonian igneous ages within the
present-day SE Appalachians [Carrigan et al., 2003] sug-
gest a complex Amazonia-Laurentia configuration. The
reconstructed orogen is marked by both strike-slip faults
and high-grade thrust faults, characteristic of a transpressive
evolution. The evolving configuration between Amazonia
and Laurentia requires fewer cratonic elements than a static,
postcollisional geometry to account for �3000 km of
Grenvillian deformation along the eastern margin of
Laurentia. Thus models which call for the involvement
of the Kalahari or Rio de Plata cratons along the Grenville
margin of Laurentia may not be justified. We note that

motion between Amazonia and Laurentia does not affect
the position of Baltica opposite the Labrador margin,
whose role in ‘‘Grenvillian’’ history and position within
Rodinia is supported on the basis of both geological
[Gower et al., 1990] and paleomagnetic observations [Weil
et al., 1998].

7. Conclusions

[30] The Nova Brasilândia belt in the SW portion of the
Amazon craton comprises deep-water sediments (turbiditic
greywackes and marls) and two intrusive granite suites. The
lithologies and structural trends of the belt are distinct from
the polymetamorphic granitoids that make up the Amazon
basement to the north. Petrological observations coupled
with geochronological data from different isotopic systems
confirm that metasedimentary rocks exposed in the Nova
Brasilândia belt equilibrated at midcrustal levels at circa
1.09 Ga. Structural data from the deformed belt suggest
that metamorphism resulted from crustal thickening in a
transpressional zone, with strain partitioned between
thrust sheets and sinistral, strike-slip faults. Regional aero-

Figure 12. Schematic diagram in map view illustrating the alternating phases of transpression and
transtension experienced by the Amazon craton in its Grenville interaction with Laurentia. Strike-slip
motion of the Amazon craton past Laurentia began with a collision with the southern portion of Laurentia
(Llano Uplift region) and proceeded with motion past the SE promontory of the North American craton,
resulting in the development of a large releasing bend in the orogenic framework. The Sunsás/Grenville
belt lies between the Paragua craton and Laurentia with the Nova Brasilândia belt separates the former
from the Amazon craton.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

18 of 20

TC6004


magnetic anomaly data and geochronological data indicate a
linear E-W trend for the NBMB extending for �2000 km,
as well as a limit to the basement rocks of the craton. These
observations are interpreted to indicate the location of the
Amazon cratonic margin during Mesoproterozoic times.
One consequence of this is that models of the Amazon
craton where basement age provinces are correlated across
the Nova Brasilândia belt [e.g., Tassinari and Macambira,
1999; Tassinari et al., 2000; Santos et al., 2000; Geraldes et
al., 2001] need to be reevaluated.
[31] We propose a new tectonic scenario for the deposi-

tion of the Nova Brasilândia sediments into a large releasing
bend of a dominantly strike-slip margin between Amazonia
and southernmost Laurentia. This deposition postdates the
original collision between these cratons at circa 1.2 Ga and
is reflective of a transtensional phase in the ongoing
northward motion of Amazonia past Laurentia. High-grade

metamorphism from crustal thickening and granite emplace-
ment marks the suturing of the Paragua block to the
(present-day) southern margin of the Amazon craton in
latest Mesoproterozoic times (circa 1.09 Ga). As such, the
Nova Brasilândia belt marks the limits of the Amazon
craton for most of the Mesoproterozoic, a fact that must
be taken into account for Rodinia reconstructions for this
time period.

[32] Acknowledgments. Funds for the completion of fieldwork
were received from a student grant from the Geological Society of America,
the Turner Fund of the University of Michigan, and the Stichting Dr.
Shurmannfonds and analytical support from the U.S. National Science
Foundation (EAR 0230059). U-Pb laboratory work in Germany was
completed with the benefit of a fellowship from the Deutscher Akade-
mischer Austauschdienst. Marcus Johnson and Chris Hall are thanked for
their help in the University of Michigan 40Ar/39Ar laboratory. Helpful
reviews by Randy Van Schmus, Kent Condie, and Associate Editor David
Evans improved the final manuscript.

References

Almeida, F. F. M., and Y. Hasui (1984), O embasamento
da plataforma Sul Americana, inOPrecambriano do

Brasil, edited by F. F. M. Almeida and Y. Hasui, Ed.
Edgard Blücher, São Paulo.

Alvarenga, C. J. S., and R. Trompette (1993), Evolução
tectônica Brasiliana da Faixa Paraguai: A estrutur-
ação da região de Cuiabá, Rev. Bras. Geocienc., 23,
18– 30.

Alvarenga, C. J. S., C. A. V. Moura, P. Gorayeb, P. S.
Sousa, and F. A. M. Abreu (2000), Paraguay and
Araguaia belts, in Tectonic Evolution of South
America: 31st International Geological Congress,
Rio de Janeiro, Brazil, August 6 –17, 2000, edited
by U. G. Cordani, E. J. Milani, and A. Thomaz
Filho, pp. 183–193, Fundo Setorial do Pet. e Gás
Nat. (FINEP), Rio de Janeiro.

Amaral, G. (1974), Geologia Pré-Cambriana da região
Amazônica, Ph.D. thesis, 144 pp., Univ. de São
Paulo, São Paulo.

Bahia, R. B. C. (1997), A Formação Palmeiral (Proter-
ozóico Superior) na Serra dos Pacaás Novos, Oeste
de Rondônia, MSc. thesis, 83 pp., Univ. Fed. Pará,
Belém, Brazil.

Bahia, R. B. C., and C. R. Silva (1998), Programa
levantamento geológicos básicos do Brasil, Rio
Pardo, Folha SC.20–Z-C-VI, 120 pp., Co. de Pes-
quisa de Recursos Minerais, Brasilia.

Barros, A. M., R. H. Silva, O. R. F. Cardoso, F. A.
Freire, J. J. Souza, M. Rivetti, D. S. Luz, R. C.
Barros Palmeira, and C. C. G. Tassinari (1978),
Folha SD.21 Cuiabá, Projeto RADAMBRASIL,
vol. 26, pp. 25 – 176, Dep. Nac. de Producao
Mineral, Rio de Janeiro.

Bettencourt, J. S., T. C. Onstott, T. de Jesus, and
W. Teixeira (1996), Tectonic interpretation of
40Ar/39Ar ages of country rocks from the central
sector of the Rı́o Negro-Juruena Province, south-
west Amazonian craton, Int. Geol. Rev., 38, 42 –56.

Bettencourt, J. S., R. M. Tosdal, W. B. Leite, and B. L.
Payolla (1999), Mesoproterozoic rapakivi granites
of the Rondônia tin province, southwestern border
of the Amazonian Craton, Brazil: Reconnaissance
U-Pb geochronology and regional implications,
Precambrian Res., 95, 41–67.

Bohlen, S. R., and E. J. Essene (1977), Feldspar and
oxide thermometry of granulites in the Adirondack
Highlands, Contrib. Mineral. Petrol., 62, 153–169.

Bond, G. C., P. A. Nickeson, and M. A. Kominz
(1984), Breakup of a supercontinent between
625 Ma and 555 Ma; new evidence and implica-
tions for continental histories, Earth Planet. Sci.,

Lett., 70, 325–345.

Carrigan, C. W., C. F. Miller, P. D. Fullagar, B. R.
Bream, R. D. Hatcher, and C. D. Coath (2003),
Ion microprobe age and geochemistry of southern
Appalachian basement, with implications for Pro-
terozoic and Paleozoic reconstructions, Precam-
brian Res., 120, 1 –36.

Companhia de Pesquisa de Recursos Minerais (1983a),
Projeto Serra dos Parecis, in Ministério das Minas e
Energia, Departamento Nacional da Produção

Mineral, Relatório Final, vols. I and II, anexos I
and II, Rio de Janeiro. (Reproduced in Brazil Air-
borne Radiometric Mapping Project (BARMP),
technical report and survey atlas, Geol. Surv. of
Can., Ottawa, 1997.)

Companhia de Pesquisa de Recursos Minerais (1983b),
Projeto Pacaás Novos, in Ministério das Minas e

Energia, Departamento Nacional da Produção
Mineral, Relatório Final, vols. I and II, anexos I
and II, Rio de Janeiro. (Reproduced in Brazil Air-
borne Radiometric Mapping Project (BARMP),
technical report and survey atlas, Geol. Surv. of
Can., Ottawa, 1997.)

Companhia de Pesquisa de Recursos Minerais (1983c),
Projeto Rio do Sangue, in Ministério das Minas e
Energia, Departamento Nacional da Produção
Mineral, Relatório Final, vols. I and II, anexos I
and II, Rio de Janeiro. (Reproduced in Brazil Air-
borne Radiometric Mapping Project (BARMP),
technical report and survey atlas, Geol. Surv. of
Can., Ottawa, 1997.)

Cosca, M. A., J. F. Sutter, and E. J. Essene (1991),
Cooling an inferred uplift/erosion of the Grenville
orogen, Ontario: Constraints from 40Ar/39Ar geo-
chronology, Tectonics, 10, 959–977.

Dalziel, I. W. D. (1991), Pacific margins of Laurentia
and East Antarctica-Australia as a conjugate rift
pair: Evidence and implications for an Eocambrian
supercontinent, Geology, 19, 598–601.

Davidson, A. (1998), An overview of the Grenville
Province, Canadian Shield, in Geology of the Pre-
cambrian Superior and Grenville Provinces and

Precambrian Fossils in North America, Geol.
Can., vol. 7, pp. 205 –270, Geol. Surv. of Can.,
Ottowa.

Dutrow, B. L., and M. J. Holdaway (1989), Experimen-
tal determination of the upper thermal stability of
Fe-staurolite + quartz at medium pressures, J. Pet-
rol., 30, 229–248.

ENCAL (1989), Projeto Aeromagnetométrico Bacia
dos Parecis, Relatório Final de Aquisição e Proces-
samento de Dados, vol. 9, Pet. Bras., Rio de
Janeiro.

England, P. C., and A. B. Thompson (1984), Pressure-
temperature-time paths of regional metamorphism,
part I, Heat transfer during the evolution of regions
of thickened continental crust, J. Petrol., 25, 894–
928.

Fernandes, C. J. (1999), Geologia do depósito Pau-a-
Piques e guias prospectivos para ouro no Grupo
Aguapeı́, sudoeste do estado de Mato Grosso,
MSc. thesis, 134 pp., Univ. Fed. Rio Grande do
Sul, Porto Allegre, Brazil.

Ganguly, J., W. Cheng, and M. Tirone (1996), Thermo-
dynamics of aluminosilicate garnet solution: New
experimental data, an optimized model, and ther-
mometric applications, Contrib. Mineral. Petrol.,
126, 137–151.

GEOMAG (1995), Prospecções Aerogeofı́sicas, Le-
vantamento Aeromagnetométrico e Aerogravimé-
trico na Bacia dos Parecis (Sub-Bacia do Alto
Xingu)—Equipe EAG-01, Relatório Final de
Aquisição e de Processamento de Dados, Pet.
Bras., vol. 16, Rio de Janeiro.

GEOMAG (1996), Prospecções Aerogeofı́sicas, Le-
vantamento Aeromagnetométrico e Aerogravimé-
trico na Reserva Indı́gena Juruena e Nordeste do
Campo Novo, Relatório Final de Aquisição e de
Processamento de Dados, Pet. Bras., vol. 16, Rio
de Janeiro.

Geraldes, M. C., B. R. Figueiredo, C. C. G. Tassinari,
and H. D. Ebert (1997), Middle Proterozoic vein
hosted gold deposits in the Pontes e Lacerda region,
southwestern Amazonian craton, Brazil, Int. Geol.
Rev., 39, 438–448.

Geraldes, M. C., W. R. Van Schmus, K. C. Condie,
S. Bell, W. Teixeira, M. Babinski, J. K. Bartley,
and L. C. Kah (2001), Proterozoic geologic evolu-
tion of the SW part of the Amazonian Craton in
Mato Grosso State, Brazil, Precambrian Res., 111,
91 –128.

Gower, C. F., A. B. Ryan, and T. Rivers (1990), Mid-
Proterozoic Laurentia-Baltica: An overview of its
geological evolution and a summary of the contri-
butions made by this volume, in Mid-Proterozoic
Laurentia-Baltica, edited by C. F. Gower, T. Rivers,
and A. B. Ryan, Geol. Assoc. Can. Spec. Pap., 38,
1 –20.

Guidotti, C. V., and M. D. Dyar (1991), Ferric iron in
metamorphic biotite and its petrologic and crystal-
lochemical implications, Am. Mineral., 76, 161 –
175.

Hoffman, P. F. (1991), Did the breakout of Laurentia
turn Gondwana inside out?, Science, 252, 1409–
1412.

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

19 of 20

TC6004


Holland, T. J. B., and R. Powell (1998), An internally
consistent thermodynamic dataset for phases of pet-
rologic interest, J. Metamorph. Geol., 16, 309 –
343.

Krogh, T. E. (1994), Precise U-Pb ages for Grenvillian
and pre-Grenvillian thrusting of Proterozoic and
Archean metamorphic assemblages in the Grenville
Front Tectonic Zone, Canada, Tectonics, 13, 963–
982.

Leal, J. W. L., G. H. Silva, D. B. Santos, W. Teixeira,
M. I. C. de Lima, C. A. C. Fernandes, and A. Pinto
(1978), Folha SC.20, Porto Velho, Projeto
RADAMBRASIL, vol. 16, pp. 17 –184, Dep. Nac.
de Producao Miner., Rio de Janeiro.

Litherland, M., and G. Power (1989), The geologic and
geomorphologic evolution of Serrania Huanchaca,
eastern Bolivia: The legendary ‘‘Lost World’’, J. S.
Am. Earth Sci., 2, 1 – 17.

Litherland, M., et al. (1986), The Geology and Mineral
Resources of the Bolivian Precambrian Shield,
Overseas Mem., vol. 9, 153 pp., Br. Geol. Surv.,
London.

Litherland, M., et al. (1989), The Proterozoic of eastern
Bolivia and its relationship to the Andean mobile
belt, Precambrian Res., 43, 157–174.

Luft, J. L., Jr., G. Rizzotto, F. Chemale Jr., and E. F.
Lima (2000), Análise geológica-estrutural do cin-
turão Sunsás na região de Nova Brasilândia, sudeste
de Rondônia, Rev. Pesq. Geocienc., 2, 65–78.

Mezger, K., B. A. van der Pluijm, E. J. Essene, and
A. N. Halliday (1992), The Carthage-Colton mylo-
nite zone (Adirondack Mountains, New York): The
site of a cryptic suture in the Grenville Orogen?,
J. Geol., 100, 630–638.

Moecher, D. P., E. J. Essene, and L. M. Anovitz (1988),
Calculation and application of clinopyroxene-
garnet-plagioclase-quartz geobarometers, Contrib.
Mineral. Petrol., 100, 92–106.

Montes-Lauar, C. R., I. G. Pacca, A. J. Melfi, E. M.
Piccirillo, G. Bellieni, R. Petrini, and R. Rizzieri
(1994), The Anari and Tapirapua Jurassic forma-
tions, western Brazil: Paleomagnetism, geochemis-
try and geochronology, Earth Planet. Sci. Lett.,
128, 357–371.

Mosher, S. (1998), Tectonic evolution of the southern
Laurentian Grenville orogenic belt, Geol. Soc. Am.
Bull., 110, 1357–1375.

Payolla, B. L., J. S. Bettencourt, M. Kozuch, W. B.
Leite, A. H. Fetter, and W. R. Van Schmus
(2002), Geological evolution of the basement rocks
in the east-central part of the Rondônia Tin
Province, SW Amazonian craton, Brazil: U-Pb
and Sm-Nd isotopic constraints, Precambrian

Res., 119, 141–169.
PROSPEC (1978), Geologia, Prospecções E Aerofoto-

grametria, Projeto Cabeceiras do Rio Guaporé,

relatório final, texto e anexos, Minist. das Minas e
Energia, Dep. Nac. da Produção Mineral, Co. de
Pesquisa de Recursos Minerais, Rio de Janeiro.
(Reproduced in Brazil Airborne Radiometric Map-
ping Project (BARMP), technical report and survey
atlas, Geol. Surv. of Can., Ottawa, 1997.)

Rizzotto, G. J. (1999), Petrologia e geotectônia do
Grupo Nova Brasilândia, Rondônia, MSc. thesis,
131 pp., Univ. Fed. Rio Grande do Sul, Porto
Allegre, Brazil.

Rizzotto, G. J. (2001), Reavaliação do ciclo orogênico
Sunsás/Aguapeı́ no sudoeste do craton Amazônico,
paper presented at Workshop of the SWAmazonian
Craton: State of the Art, Univ. de São Paulo, São
Paulo, Brazil, August 10 –12.

Rizzotto, G. J., F. Chemale, E. F. de Lima, R. Van
Schmus, and A. Fetter (1999), Sm/Nd and U/Pb
isotopic data for the Nova Brasilândia metaplutonic,
metavolcanosedimentary sequence, Rondônia, in
paper presented at the 70th Meeting of the Brazilian
Geological Society, Salvador, Brazil.

Sadowski, G. R., and J. S. Bettencourt (1996), Meso-
proterozoic tectonic correlations between eastern
Laurentia and the western border of the Amazon
craton, Precambrian Res., 76, 213 –227.

Saes, G. S. (1999), Tectonic and paleogeographic evo-
lution of the Aguapeı́ aulacogen (1.2 – 1.0 Ga) and
the basement terranes in the southern Amazon cra-
ton, Ph.D. thesis, 135 pp., Univ. de São Paulo, São
Paulo.

Santos, J. O. E., L. A. Hartmann, H. E. Gaudette, D. I.
Groves, N. J. Mcnaughton, and I. R. Fletcher
(2000), A new understanding of the provinces of
the Amazon craton based integration of field
mapping and U-Pb and Sm-Nd geochronology,
Gondwana Res., 3, 454–488.

Santos, J. O. E., G. A. Rizzotto, L. A. Hartmann, N. J.
McNaughton, and I. R. Fletcher (2001), Ages of
sedimentary basins related to the Sunsás and
Juruena orogenies, southwest Amazon craton estab-
lished by zircon U-Pb geochronology, paper pre-
sented at Workshop of the SW Amazonian Craton:
State of the Art, Univ. de São Paolo, São Paulo,
Brazil, August 10–12.

Scandolara, J. E., and G. J. Rizzotto (1998), Pro-
grama Levantamento Geológicos do Brasil,
Paulo Saldanha, Folha SC.20 –Z-C-V, Estado
de Rondônia –Escala 1:100,000, 140 pp., Co.
de Pesquisa de Recursos Minerais, Brasilia.

Scandolâra, J. E., G. J. Rizzotto, J. L. de Amorim, R. B.
C. Bahia, M. L. Quadros, and C. R. da Silva (1999),
Geological map of Rondônia 1:1,000,000. Co. de
Pesquisa de Recursos Minerais, Brasilia.

Svisero, D. P. (1995), Distribution and origin of
diamonds in Brazil: An overview, J. Geodyn.,
20, 493–514.

Tassinari, C. C. G., and M. J. B. Macambira (1999),
Geochronological provinces of the Amazonian
craton, Episodes, 22, 174 –182.

Tassinari, C. C. G., J. S. Bettencourt, M. C. Geraldes,
M. J. B. Macambria, and J. M. Lafon (2000), The
Amazonian craton, in Tectonic Evolution of South
America: 31st International Geological Congress,

Rio de Janeiro, Brazil, August 6–17, 2000, edited
by E. J. Milani and A. Thomaz Filho, pp. 41–95,
Fundo Setorial do Pet. e Gás Nat. (FINEP), Rio de
Janeiro.

Thomas, W. A., and R. A. Astini (1996), The Argentine
Precordillera: A traveler from the Ouachita embay-
ment of North American Laurentia, Science, 273,
752 –757.

Thompson, A. B. (1976), Mineral reactions in politic
rocks 2, calculations of some P-T-X (Fe-Mg) phase
relations, Am. J. Sci., 276, 425–454.

Tohver, E., B. A. van der Pluijm, R. Van der Voo,
G. Rizzotto, and J. E. Scandolara (2002), Paleo-
geography of the Amazon craton at 1.2 Ga: Early
Grenvillian collision with the Llano segment of
Laurentia, Earth Planet. Sci. Lett., 199, 185 –
2000.

Valley, J. W. (1992), Granulite processes driven by
magmatic processes in the deep crust, Earth Sci.
Rev., 32, 145–146.

Weil, A. B., R. Van der Voo, C. Mac-Niocaill, and J. G.
Meert (1998), The Proterozoic supercontinent
Rodinia: Paleomagnetically derived reconstruction
for 1100 to 800 Ma, Earth Planet. Sci. Lett., 154,
13 –24.

Wilding, M. C., B. Harte, and J. W. Harris (1991),
Evidence for a deep origin of São Luis diamonds,
in Extended Abstracts of the 5th International

Kimberlite Congress, Spec. Publ., 2/91, pp.
456 –458, Co. de Pesquisa de Recursos Minerais,
Brasilia.

���������
E. Essene and B. van der Pluijm, Department of

Geological Sciences, University of Michigan, 2534
C. C. Little Building, Ann Arbor, MI 48109-1063, USA.

K. Mezger, Institut für Mineralogie, Universität
Münster, Corrensstrasse. 24, D-48149 Münster,
Germany.

G. Rizzotto and J. Scandolara, Companhia de
Pesquisas de Recursos Minerais (CPRM), Av. Lauro
Sodré 2561, Porto Velho, RO, Brazil 78904-300.

E. Tohver, Instituto de Geociências (GMG),
Universidade de São Paulo, Rua do Lago, 562, Cidade
Universitária, 05508-900 São Paulo, SP Brazil.
(etohver@usp.br)

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON

20 of 20

TC6004


Figure 11. Aeromagnetic anomaly data for the SW Amazon craton. The Nova Brasilândia trend is
marked by a clear 1000 km long anomaly that is covered in the eastern extent by the Paleozoic sediments
of the Pimenta Bueno graben (labeled Pz) and Cretaceous sediments of the Parecis Formation (labeled
Kt). Locations of known kimberlite pipes are designated with white stars [Companhia de Pesquisa de
Recursos Minerais, 1983a, 1983b, 1983c; ENCAL, 1989; GEOMAG, 1995, 1996; PROSPEC, 1978].
(ENCAL report is archived in SEDOC/PETROBRAS headquarters in Rio de Janeiro under registration
number 103–07407; GEOMAG [1995] report is archived in the SEDOC/GEREX/E&P headquarters in
Rio de Janeiro under registration number 101–09824; GEOMAG [1996] report is archived in the
SEDOC/GEREX/E&P headquarters in Rio de Janeiro under registration number 101–09974.)

TC6004 TOHVER ET AL.: MESOPROTEROZOIC MARGIN OF AMAZON CRATON TC6004

17 of 20