Thermochronology, geochronology, and upper crustal structure of the Cordillera Real: Implications for Cenozoic exhumation of the central Andean plateau

Abstract

1] Structural mapping, 40 Ar/ 39 Ar and fission track thermochronology, U-Pb geochronology, and basin analysis reveal rapid cooling during middle Eocene– late Oligocene and late Miocene–Pliocene exhumation in the central Andean plateau of Bolivia. In the 4– 6 km high Cordillera Real, numerous granites and SW directed fold-thrust structures define the central Andean backthrust belt along the Altiplano–Eastern Cordillera boundary. U-Pb zircon analyses indicate Permo-Triassic granitic magmatism, with less extensive magmatism of late Oligocene age. Mapping reveals low magnitudes of slip (<2–5 km) for most faults on the basis of unit thicknesses, stratigraphic separation, and cutoff relationships. These results suggest that a deeper structure was probably involved in exhumation of rocks from >5 km depth. The $26 Ma Quimsa Cruz granite postdated most thrust structures, suggesting that upper crustal shortening in the Cordillera Real had largely ceased by late Oligocene time. Results of 40 Ar/ 39 Ar and fission track modeling help constrain the moderate to low-temperature (<350°C) cooling history, revealing two phases of rapid cooling from 45–40 Ma to 26 Ma and from $11 Ma onward. Initial cooling coincided with middle Eocene–late Oligocene deformation in the backthrust belt and associated deposition of coarse clastic sediments in the Altiplano basin. Eocene-Oligocene exhumation of $7.5 km of upper crust is estimated on the basis of thermochronologic data. Rapid late Miocene and younger cooling involved an estimated $3.5 km of exhumation and occurred in the apparent absence of upper crustal shortening. These findings suggest that crustal shortening and resultant exhumation of middle Eocene–late Oligocene age played a major role in construction of the central Andes. However, for late Miocene exhumation, the importance of alternative, nonshortening mechanisms is difficult to ascertain due to a poor understanding of subsurface structures. We speculate that greater precipitation on the eastern edge of the central Andean plateau north of $17.5°S was a key factor in driving rapid, youthful exhumation of the Cordillera Real. Citation: Gillis, R. J., B. K. Horton, and M. Grove (2006), Thermochronology, geochronology, and upper crustal structure of the Cordillera Real: Implications for Cenozoic exhumation of the central Andean plateau, Tectonics, 25, TC6007, doi:10.1029/2005TC001887.

Full text

Thermochronology, geochronology, and upper crustal structure
of the Cordillera Real: Implications for Cenozoic exhumation
of the central Andean plateau
Robert J. Gillis,
1,2
Brian K. Horton,
1,3
and Marty Grove
1
Received 14 July 2005; revised 8 January 2006; accepted 26 July 2006; published 21 December 2006.
[1] Structural mapping,
40
Ar/
39
Ar and fission track
thermochronology, U-Pb geochronology, and basin
analysis reveal rapid cooling during middle Eocene –
late Oligocene and late Miocene–Pliocene exhumation
in the central Andean plateau of Bolivia. In the 4–
6 km high Cordillera Real, numerous granites and
SW directed fold-thrust structures define the central
Andean backthrust belt along the Altiplano–Eastern
Cordillera boundary. U-Pb zircon analyses indicate
Permo-Triassic granitic magmatism, with less extensive
magmatism of late Oligocene age. Mapping reveals
low magnitudes of slip (<2–5 km) for most faults on
the basis of unit thicknesses, stratigraphic separation,
and cutoff relationships. These results suggest that a
deeper structure was probably involved in exhumation
of rocks from >5 km depth. The 26 Ma Quimsa
Cruz granite postdated most thrust structures,
suggesting that upper crustal shortening in the
Cordillera Real had largely ceased by late Oligocene
time. Results of
40
Ar/
39
Ar and fission track modeling
help constrain the moderate to low-temperature
(<350°C) cooling history, revealing two phases of
rapid cooling from 45–40 Ma to 26 Ma and from
11 Ma onward. Initial cooling coincided with
middle Eocene–late Oligocene deformation in the
backthrust belt and associated deposition of coarse
clastic sediments in the Altiplano basin. Eocene-
Oligocene exhumation of 7.5 km of upper crust is
estimated on the basis of thermochronologic data.
Rapid late Miocene and younger cooling involved an
estimated 3.5 km of exhumation and occurred in the
apparent absence of upper crustal shortening. These
findings suggest that crustal shortening and resultant
exhumation of middle Eocene–late Oligocene age
played a major role in construction of the central
Andes. However, for late Miocene exhumation, the
importance of alternative, nonshortening mechanisms
is difficult to ascertain due to a poor understanding of
subsurface structures. We speculate that greater
precipitation on the eastern edge of the central
Andean plateau north of 17.5°S was a key factor in
driving rapid, youthful exhumation of the Cordillera
Real. Citation: Gillis, R. J., B. K. Horton, and M. Grove
(2006), Thermochronology, geochronology, and upper crustal
structure of the Cordillera Real: Implications for Cenozoic
exhumation of the central Andean plateau, Tectonics,25,
TC6007, doi:10.1029/2005TC001887.
1. Introduction
[2] The Cordillera Real forms high topography in one of
the most intensely shortened and thickened region of the
Andes [Dorbath et al., 1993; Baby et al., 1997; Beck and
Zandt, 2002]. At 4–6.4 km altitude, the range marks both
the eastern topographic margin of the central Andean
plateau, defined as the low-relief, generally internally
drained region above 3 km [Isacks, 1988], and the structural
boundary between the Altiplano and Eastern Cordillera
provinces (Figure 1). As the principal divide separating
the closed Altiplano basin and Amazon drainage system, the
Cordillera Real has profoundly influenced orographic pre-
cipitation, regional climate, and geomorphic evolution of
the central Andes [Masek et al., 1994; Horton, 1999;
Montgomery et al., 2001]. The range also occupies the
transition from hinterland- to foreland-directed thrust struc-
tures [McQuarrie and DeCelles, 2001; McQuarrie, 2002]
and coincides with a lithospheric boundary possibly repre-
senting an ancient suture zone [Dorbath et al.,1993].
Furthermore, uplift of the Cordillera Real presumably
helped dictate the Cenozoic sedimentation history of the
Altiplano [Sempere et al., 1990; Horton et al., 2001, 2002].
[3] Despite its fundamental role in construction of the
central Andean plateau, the timing, magnitude, and style of
deformation, and associated exhumation and basin devel-
opment remain poorly understood. Multiple granitic bodies
compose the Cordillera Real, but disagreement persists over
emplacement ages and subsequent cooling histories [e.g.,
McBride et al., 1983; Miller, 1988; Miller and Harris, 1989;
Lamb and Hoke, 1997]. Previous thermochronologic
results, notably the fission track data of Benjamin et al.
[1987], have been variably interpreted to suggest an Eocene
TECTONICS, VOL. 25, TC6007, doi:10.1029/2005TC001887, 2006
Click
Here
for
Full
A
rticl
e
1
Department of Earth and Space Sciences, University of California, Los
Angeles, California, USA.
2
Now at the Alaska Division of Geological and Geophysical Surveys,
Fairbanks, Alaska, USA.
3
Now at the Department of Geological Sciences and Institute for
Geophysics, Jackson School of Geosciences, University of Texas at Austin,
Austin, Texas, USA.
Copyright 2006 by the American Geophysical Union.
0278-7407/06/2005TC001887$12.00
TC6007 1of22

Figure 1
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
2of22
TC6007

pulse of heating, Eocene pulse of cooling, uniform Eocene
to modern cooling, Miocene linear or stepwise increase in
cooling, and late Miocene exponential increase in cooling
[Benjamin, 1986; Benjamin et al., 1987; Masek et al., 1994;
Lamb and Hoke, 1997; Safran, 1998; Moore and England,
2001; Anders et al., 2002]. These conflicting interpretations
demonstrate the large uncertainties in the exhumation his-
tory of the Cordillera Real. An accurate record is essential
for understanding mountain building and plateau evolution
in the central Andes, and the interactions among tectonics,
erosion, and climate.
[4] The purpose of this paper is to evaluate the exhumation
record of the Cordillera Real along three mapping traverses at
16–17°S. Widespread granite exposures afford opportunities
to assess the range’s cooling history and crosscutting relation-
ships with newly identified structures. U-Pb zircon analyses
constrain Permo-Triassic and late Oligocene granite emplace-
ment ages, and
40
Ar/
39
Ar, apatite fission track, and applicable
modeling results provide a record of exhumation-induced
cooling. These results indicate rapid cooling during middle
Eocene to early Miocene time, considerably earlier than
generally assumed. Exhumation of this age is consistent with
sediment provenance data for clastic fill in the northern
Altiplanobasindatedbetween34and24Mabynew
40
Ar/
39
Ar results for interbedded tuffs. Collectively, these
findings require revision of previous estimates for the timing
of initial topographic expression and tempo of exhumation
along the eastern margin of the central Andean plateau.
2. Geologic Setting
[5] The Cordillera Real is a 30 250 km range along the
eastern flank of the central Andean plateau (Figure 1). This
rugged, glaciated range represents the highest nonvolcanic
topography in Bolivia, with average elevations of 5km
and major peak elevations of 6.1 –6.4 km. Although it forms
the NW trending divide between the Amazon drainage and
internally drained Altiplano, the range has been breached
locally by headward erosion of the Rio La Paz, a tributary of
the Amazon. The entire range is recognized as the Cordil-
lera Real, but high topography southeast of the Rio La Paz
is commonly identified separately as the Quimsa Cruz
(Figure 1a).
[6] The Cordillera Real lies within the Huarina fold-
thrust belt (HFTB) of the westernmost Eastern Cordillera
(Figure 1). The HFTB represents part of the W to SW
directed central Andean backthrust belt that extends from 15
to 22°S in the hinterland of the central Andes [Newell, 1949;
Pareja et al., 1978; Martinez, 1980; Roeder, 1988; Sempere
et al., 1990; McQuarrie and DeCelles, 2001]. Ordovician
strata are the structurally lowest rocks exposed and define
an axis between a NE directed thrust system and the SW
directed backthrust belt. Younger rocks in the HFTB include
Upper Devonian with limited Carboniferous and Cretaceous
rocks. Frontal thrusts of the HFTB, including the SW
directed Huarina and Coniri faults (Figure 1b), place Paleo-
zoic rocks on mainly mid-Cenozoic strata, but are com-
monly covered by younger Altiplano fill [Sempere et al.,
1990; Marsh et al., 1992].
[7] In this study, three NE-SW transects were mapped at
1:50,000 scale to provide shortening estimates and a struc-
tural framework for interpreting thermochronologic results.
The 20–40 km long transects are orthogonal to structural
trends and continue northeastward from the Altiplano mar-
gin. The northern (Amaguaya), central (Rio Zongo), and
southern (Quimsa Cruz) traverses contain similar structures
and extensive granites that intrude mainly Ordovician-
Silurian rocks (Figures 2 and 3). Low-grade metamorphic
conditions affected Paleozoic strata, producing slate, phyl-
lite, and minor quartzite [Martinez, 1980]. The main strati-
graphic units include: Upper Ordovician Coroico (2300 m)
and Amutara (800–3300 m) formations; Silurian
Cancan˜iri (100–335 m), Uncia (900 – 1300 m), and
Catavi (480–600 m) formations; Devonian Vila Vila For-
mation (700–850 m); and local occurrences of Carbonif-
erous, Permian, and Cretaceous units [Servicio Geolo´gico de
Bolivia (Geobol), 1993, 1995, 1997; McQuarrie and
DeCelles, 2001]. Major de´collement horizons are in the
Coroico, Uncia, and Cancan˜iri formations.
[8] Granites in the Cordillera Real are of Permo-Triassic
and mid-Cenozoic age. In the north, four Permo-Triassic
intrusions occupy the highest parts of the range: the Illampu
and Yani plutons in the northern transect and the Huayna
Potosı´ and Zongo plutons in the central transect (Figure 2).
Although not continuously exposed, these mineralogically
similar two-mica granites may be genetically related
[McBride et al., 1983]. Whereas the Illampu and Huayna
Potosı´ granites are generally nonfoliated and medium-
grained, the Zongo granite is much coarser with a pervasive
NE dipping foliation; foliation in the medium-grained Yani
granite only occurs locally along its NE margin. In the south,
the Quimsa Cruz pluton cuts Paleozoic rocks (Figure 3) and
is composed of a nonfoliated granodiorite and a porphyritic
monzogranite.
3. Structural Geology
[9] Three mapping transects (Figures 1– 3) were selected
to maximize exposure, relief, and accessibility. Most struc-
tures fall into three categories: (1) NE dipping faults cutting
Paleozoic strata with hanging wall on footwall thrust cutoff
geometries ranging from ramp on ramp to flat on flat; (2)
steeply NE dipping faults cutting granites with possible
hanging wall ramp on footwall ramp thrust relationships;
and (3) fault-bend, fault propagation, and buckle folds in
Paleozoic strata. Tentative fault slip estimates are made on
Figure 1. (a) Regional topography of the central Andes showing boundaries (dashed lines) between major
tectonomorphic zones [after McQuarrie et al., 2005a]. Darker colors define high topography (>4.5 km) of Cordillera
Real. (b) Regional geologic map of the Cordillera Real and surrounding regions of the central Andean backthrust belt along
the Eastern Cordillera–Altiplano boundary. Areas of white shading show extent of three mapping transects. CRFZ,
Cordillera Real fault zone.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
3of22
TC6007

Figure 2
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
4of22
TC6007

the basis of stratigraphic separation, unit thicknesses, fault
cutoff relationships, and inferred geometries at depth. Cross
sections (Figure 4) are not balanced and are presented only
as representations compatible with mapped surface struc-
tures; alternative interpretations are permissible.
3.1. Northern Transect: Amaguaya Region
[10] Permo-Triassic granites intrude the Ordovician
Amutara Formation (Oam) and are cut by steeply NE
dipping thrusts along the northern transect. The moderately
SW dipping Amaguaya thrust marks the northern edge of
the transect (Figures 2 and 4a). A prominent fault in the
middle of the transect places the Yani pluton over the
Illampu pluton; regionally, this fault separates gently dip-
ping Oam strata and low-amplitude fold trains in the NE
from steeply NE dipping Oam strata in the SW (Figure 2).
The southern edge of the transect consists of folded and
thrusted slates of the Silurian Cancan˜iri Formation (Scc)
along the eastern margin of the Altiplano. The lack of
cutoffs within granites and stratigraphic uncertainties pre-
clude shortening estimates for this transect.
3.1.1. NE Part: Amaguaya Thrust
[11] The principal NE directed thrust in the region cuts
the NE margin of the Yani pluton (AA, Figure 2). This fault,
here named the Amaguaya thrust, is defined by a SW
dipping (49 – 54°)10–20 cm wide ductile shear zone
enveloped within an 200 m wide brittle shear zone. In
both hanging wall and footwall, pervasive fractures and a
well-developed foliation in the Yani granite dip 20– 70°SW.
Shear sense indicators and slickenlines (stereonets, Figure 2)
indicate top-to-NE transport. In map view, the fault trace
projects to the ESE into shallowly dipping Ordovician
Amutara (Oam) strata that occur in both hanging wall and
footwall, suggesting limited stratigraphic separation. A lack
of Oam marker horizons precludes slip estimates.
3.1.2. ENE Part: Ordovician Rocks
[12] The ENE part of the transect is dominated by
shallowly dipping Oam strata. Oam is in intrusive contact
with the Yani granite (AB, Figure 2) except near the NE
edge of the transect where it is in fault contact along a SW
dipping thrust fault. The intrusive Oam-Yani contact paral-
lels bedding and dips 4°NE over a 10 km distance (AB
and AC). Low-amplitude NW trending folds are common in
Oam strata to the NE (AD). In the SSE, the Yani granite
intruded gently folded Oam strata (AC and AE) considered
to be the roof of the Yani pluton (AF). A potential Yani roof
contact (AB) is truncated by a NE dipping fault (AG) with
only 60 m of throw.
Figure 2. Geologic maps of northern (Amaguaya) and central (Rio Zongo) transects. Location is shown in Figure 1b.
Black circles identify granite samples for thermochronologic studies. Lettered rectangles (AA –AM; ZA– ZM) identify
locations discussed in text. Stereonets represent fault orientation and slickenline data.
Figure 3. Geologic map of southern (Quimsa Cruz) transect. Location is shown in Figure 1b. Black
circles identify granite samples for thermochronologic studies. Lettered rectangles (QA – QE) identify
locations discussed in text. Rose diagram represents joint and fracture orientations.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
5of22
TC6007

3.1.3. Middle Part: Yani-Illampu Thrust
[13] A major NE dipping (63°) fault juxtaposes the
Yani and Illampu plutons (AH, Figure 2). This structure, the
Yani-Illampu thrust, continues >170 km to the NW and
places Ordovician over Devonian rocks [Servicio Nacional
de Geologı´a y Te´cnico de Minas and Yacimientos
Petrolı´feros Fiscales Bolivianos, 2003], a minimum strati-
graphic separation of 5 km. Displacement decreases to the
SE, with the Zongo granite and Oam strata thrust onto Oam
rocksinahangingwallflatonfootwallrampcutoff
relationship in the central transect (ZD, Figure 2).
3.1.4. SW Margin of Illampu Granite
[14] The SW margin of the Illampu pluton is generally in
fault contact with steeply NE dipping Oam strata. A NE
dipping (75°) strand of this fault (AI, Figure 2) is defined
by fracture and gouge zones exhibiting slickensides of
variable orientation. To the NW, the fault has been displaced
1 km by a NE striking left-lateral fault (AJ), consistent
with centimeter-scale left-lateral shear indicators (AK) and
NE striking subvertical fractures with gently NE plunging
lineations (AL).
3.1.5. SW Part: Paleozoic Rocks
[15] Steeply NE dipping strata in the SW are potentially
overturned near the Ordovician-Silurian contact (AM,
Figure 2), although mud-draped ripple marks in the Silurian
Cancan˜iri Formation (Scc) show upright orientations locally.
A proposed fault placing overturned, steeply NE dipping
Scc hanging wall rocks over relatively flat-lying Scc strata
accommodated an unknown amount of slip.
3.2. Central Transect: Rio Zongo Region
[16] Mapping of a series of NE dipping thrusts and the
Permo-Triassic Zongo and Huayna Potosı´ granites along the
central transect helps place thermochronologic data [e.g.,
McBride et al., 1983, 1987; Benjamin et al., 1987] into a
structural context (Figure 2 and 4b). Ordovician-Silurian
strata and the two plutons are well exposed at high eleva-
tion, but mapping in more vegetated areas below 3.5 km
was performed primarily by roadcut inspection along the
Zongo valley. Here, faults were identified locally and
tentatively projected along strike using measured orienta-
tions, topographic features, and satellite imagery.
3.2.1. NE Part: Zongo Granite
[17] In the NE, the Zongo pluton, Ordovician Coroico
(Ocr), and Amutara (Oam) formations are deformed by
SW directed thrusts. Displacement is difficult to con-
strain, but unit thicknesses indicate <2.5 km of strati-
graphic separation for faults placing Ocr on footwall Ocr
and Oam strata (ZA, Figure 2) [Geobol,1995].The
Zongo granite intruded Oam rocks along its NE and
SW margins [Heinrich, 1988] and is bisected by a NE
dipping (63°) fault (ZB) that produced minimal offset of
the SE pluton margin (ZC). Along its SW margin, the
Zongo granite is cut by a NE dipping fault (ZD) that
correlates with the Yani-Illampu thrust to the north (AH,
Figure 4. Schematic cross sections of mapped transects (locations shown in Figures 2 and 3).
(a) Northern (Amaguaya) transect. (b) Central (Rio Zongo) transect. (c) Southern (Quimsa Cruz) transect.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
6of22
TC6007

Figure 2). The SE continuation of this fault is unclear,
but probably links with a fault in Oam strata (ZE).
3.2.2. Middle Part: Paleozoic Rocks
[18] South of the Zongo pluton, a SW directed thrust places
an Oam hanging wall flat on an Oam footwall ramp (ZE,
Figure 2). Hanging wall klippen are preserved 3 km SSW of
the main fault trace (ZF). Mapped traces of the klippen are
estimated on the basis of bedding-parallel fractures and
bedding discordances observed from neighboring ridges.
Correlation of the main fault trace and klippen suggests a
NE to SW change in the footwall from a 30°N dipping ramp to
an 11°N dipping flat. These relationships reveal a minimum
displacement of 6.5 km, suggesting that this fault is one of
the larger magnitude structures of the region.
3.2.3. NE Margin of Huayna Potosı´ Granite
[19] The trace of the northern contact of the Huayna
Potosı´ pluton across the Zongo valley and flanking ridges
(ZG, Figure 2) is consistent with a N dipping contact.
However, a localized <10 m wide fracture between granite
and Oam strata dips 65°SE with E trending striae,
inconsistent with the mapped geometry. Additionally, both
the granite and Oam rocks occur locally on both sides of the
fracture. Therefore the N margin of the Huayna Potosı´
granite is not considered a major structure and is tentatively
interpreted to be a modified intrusive contact with limited
thrust or strike-slip displacement.
3.2.4. SW Margin of Huayna Potosı´ Granite
[20] Along its SW edge, the Huayna Potosı´ granite sits in
the hanging wall of a low-angle (13°), NE dipping fault that
cuts NE dipping Oam strata (ZH, Figure 2). NE plunging
fault striae indicate transport to the SW. The fault cuts
downsection to the SW across steeply dipping footwall
strata, suggesting out-of-sequence motion. In map view,
the fault defines a small salient (ZH), requiring >1 km of
heave. A structural window reveals the NE continuation of
the thrust (ZI), exposing a transition to a NE dipping (54°)
footwall ramp where the Huayna Potosı´ granite occurs in
both hanging wall and footwall.
3.2.5. Ordovician-Silurian Contact
[21] Southwest of the Huayna Potosı´ granite, structures
near the Ordovician-Silurian contact include a SW vergent,
thrust-cored anticline with steeply NE dipping Oam strata on
the NE limb. On the SW limb, Oam rocks are locally over-
turned (ZJ, Figure 2) with overlying, SW dipping Scc strata
commonly exhibiting hematitic mineralization along the
basal contact (ZK). The anticline dies out along strike to the
NW (ZJ). Kinematic indicators below the mineralized contact
include meter-scale, NE directed, ramp and flat fault geom-
etries and top-to-NE folds (ZK). Mineralization and defor-
mation may be related to bedding-parallel flexural slip
associated with SW vergent folding or tectonic wedging of
competent Oam quartzite beneath the Scc shale (ZXA,
Figure 4b). The minimum shortening produced by the ex-
posed fault propagation fold is 400 m, but wedge emplace-
ment may have accommodated several additional kilometers.
3.2.6. SW Part: Paleozoic Rocks
[22] The dominant structures in the SW include two
faulted synclines containing the Silurian Catavi Formation
(Sct) in their cores (ZL and ZM, Figure 2). Because the
Silurian Uncia Formation (Sun) appears to be relatively thin
(870 m) in this area (ZXB, Figure 4b), the Uncia section
must have been thickened in the southern syncline (ZM).
Because Sun strata are tightly folded in this locality, the Sun
was likely thickened by internal deformation. Net slip in the
Silurian section is estimated at >5 km.
3.3. Southern Transect: Quimsa Cruz Region
[23] NW trending structures (Figures 3 and 4c) are con-
centrated in Paleozoic strata on the flanks of the mid-
Cenozoic Quimsa Cruz pluton. Limited deformation affected
the granite, which lacks foliation and truncates most NE
dipping thrusts and associated folds.
3.3.1. NE Part: Paleozoic Rocks
[24] NE dipping panels of imbricated Silurian Uncia
(Sun) and Catavi (Sct) strata are exposed near the NE
margin of the Quimsa Cruz pluton (QA, Figure 3). Expo-
sure is relatively poor, but a possible repetition of Sct rocks
is interpreted on the basis of measured bedding orientations
and an assumed 500 m original stratigraphic thickness. A
prominent fault exposed near the NE margin of the granite
dips NE (53°) with NE trending striae, but does not
continue into the granite. A similar relationship is observed
1 km to the east where the granite apparently cuts a NE
dipping fault within the Sct section (QB).
3.3.2. Quimsa Cruz Granite
[25] Limited deformation of the Quimsa Cruz granite was
recorded by rare, thin, discontinuous brittle shear zones that
in some cases match linear topographic features, rare
slickensides, or subvertical joints. One prominent topo-
graphic lineament tentatively interpreted as the trace of a
steeply NE dipping fault parallels regional structures and
extends across the pluton (QC, Figure 3), continuing into
Paleozoic rocks to the SE. Brittle shear indicators in the
pluton exhibit steep 45–80°NW dips with N trending
striae. Reconnaissance mapping of the SE segment of the
lineament near the SE pluton margin (near Mina Caracoles)
suggests abundant slickensides but little apparent offset of
the intrusive contact.
[26] Although there is a continuum of steeply NW to NE
dipping fractures in the pluton (rose diagram, Figure 3),
many fractures strike 125°, similar to regional structural
trends. Possible joint faces striking 045°and 075°are
expressed by uniform surfaces that form high cliffs and are
better developed along the NE pluton margin and surround-
ing Paleozoic wall rocks.
3.3.3. SW Part: Paleozoic Rocks
[27] West of the Quimsa Cruz pluton, Ordovician–Devo-
nian strata are deformed by SW directed thrusts that place
older on successively younger rocks toward the SW. The
granite cuts Ordovician Amutara (Oam) strata along the NE
limb of a NW trending syncline (QD, Figure 3). The SW
limb contains a thrust placing Oam on Sun strata in a hanging
wall flat on footwall flat geometry, requiring 3.4 km of
minimum slip. Reported unit thicknesses [Geobol, 1997]
suggest structural thickening of the Oam section, and Oam
and Sun panels are in turn thrust over the Sct section in a flat
on flat relationship that requires 2 km of minimum slip. In
the SW part of the transect, gently dipping Sun shales are
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
7of22
TC6007

inferred to structurally overlie E dipping Sct and Devonian
Vila Vila strata in an apparent hanging wall ramp on footwall
flat relationship. The minimum estimated slip required to
produce this geometry is 2.7 km.
3.4. Summary
[28] Mapping and cross section construction reveal a
complex array of Paleozoic strata and Mesozoic-Cenozoic
granites involved in fold-thrust structures. Shear sense
indicators and slickenline data indicate principally dip slip
motion along NW striking, generally NE dipping faults
(stereonets and rose diagram, Figures 2 and 3). Stratigraphic
cutoff relationships are consistent with ramp and flat geom-
etries. Therefore most faults are interpreted as thin-skinned,
SW directed thrusts (Figure 4) developed during NE-SW
shortening, consistent with previous studies of the central
Andean backthrust belt [Newell, 1949; Pareja et al., 1978;
Martinez, 1980; Roeder,1988;Sempere et al.,1990;Geobol,
1993, 1995, 1997; McQuarrie and DeCelles,2001].
[29] Structural relationships reveal low magnitudes of
slip (<2–5 km) for most faults. Fault displacements are
poorly constrained, particularly for granites, which lack
cutoffs but were emplaced at 10 –15 km depth (3– 5 kbar)
[McBride et al., 1983; Heinrich, 1988]. Possible larger
magnitude structures potentially accommodated >5 km of
reverse slip, including from NE to SW: (1) the NE directed
Amaguaya thrust; (2) the SW directed Yani-Illampu thrust;
(3) thrusts bounding the SW margins of the Illampu and
Huayna Potosı´ granites; and (4) thrusts near the Ordovician-
Silurian contact along the SW edges of the northern and
central transects. Despite the uncertainties for individual
structures, an estimation of total displacement suggests
20 km of minimum cumulative slip, or 35% NE-SW
shortening. This minimum estimate of local strain is slightly
lower than, but compatible with, the 50% net strain
suggested by recent estimates of regional shortening
[McQuarrie, 2002; Arriagada et al., 2005].
[30] Crosscutting relationships among thrusts and gran-
ites place timing constraints on deformation. The mid-
Cenozoic Quimsa Cruz pluton truncates folds and thrusts
along its margins (QB and QD, Figure 3) and is not cut by
faults involving Permo-Triassic granites (Figure 2). The
only postulated fault that cuts the Quimsa Cruz pluton
continues into Paleozoic strata to the SE but produces
limited offset of the intrusive contact [Geobol, 1997]. These
observations indicate that most NE-SW shortening post-
dated Permo-Triassic magmatism and predated mid-Ceno-
zoic magmatism, supporting suggestions that the Quimsa
Cruz granite was emplaced after the bulk of upper crustal
shortening in the Eastern Cordillera [Evernden et al., 1977;
Lamb and Hoke, 1997].
4. U-Pb Geochronology
[31] New ion microprobe U-Pb analyses support previous
K-Ar,
40
Ar/
39
Ar, and limited conventional U-Pb measure-
ments indicating emplacement of most intrusions in the
Cordillera Real during Permo-Triassic time [Evernden et al.,
1977; McBride et al., 1983, 1987; Farrar et al., 1988;
Heinrich, 1988]. These include the Illampu (Sorata), Yani,
Huayna Potosı´ (Chucara), Zongo, and Taquesa (Mururata)
granites [Evernden et al., 1977; Martinez, 1980; McBride et
al., 1983], which are partially synchronous with Permian
granites along strike 250 km to the NW in southern Peru
[Lancelot et al., 1978; Carlier et al., 1982]. Our results also
support Cenozoic emplacement ages for at least one granite
in the region (Quimsa Cruz granite) [Evernden et al., 1977;
McBride et al., 1983; Miller, 1988; Miller and Harris, 1989;
Kennan et al., 1995]. For this granite, previous K-Ar and
40
Ar/
39
Ar biotite ages of 34.2–22.8 Ma [Evernden et al.,
1977; McBride et al., 1983; Kennan et al., 1995] and an
unpublished zircon fission track age of 38 Ma [Lamb et al.,
1997] were used to infer a crystallization age between
middle Eocene and early Miocene time.
[32] Results of U-Pb ion microprobe analyses of 78
zircons from 7 granite samples (6–18 zircons per sample)
are summarized in Figure 5 and Table 1. All uncertainties
are reported as 1sstandard errors. Complete methods and
tabulated data are presented in Table S1 (available as
auxiliary material
1
). A general issue that impacts U-Pb
analyses of the Permo-Triassic granitoids is that many
Figure 5. U-Pb concordia diagrams for analyzed zircon
grains from Cordillera Real granites. Error ellipses are
shown at 2slevel. (a) Quimsa Cruz (Mina Argentina)
granodiorite. (b) Quimsa Cruz (Mina Viloco) porphyritic
monzogranite. (c) Illampu granite. (d) Yani granite. (e)
Huayna Potosı´ granite. (f) Zongo granite.
1
Auxiliary material data sets are available at ftp://ftp.agu.org/apend/tc/
2005tc001887. Other auxiliary material files are in the HTML.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
8of22
TC6007

grains have high U contents (>1000 ppm) and low Th/U
values (<0.1), signaling high susceptibility to Pb loss and/or
metamorphic recrystallization. This possibility is supported
by the fact that U-Pb analyses obtained from the Permo-
Triassic granitoids tend to spread out over an age range of
270–220 Ma (Table S1). For the conditions under which
the analyses were undertaken, a homogeneous zircon pop-
ulation would be expected to exhibit a much smaller spread
in U-Pb age (±2% or ±5 Ma 1sat 250 Ma). Because the
observed age variability is much higher, we regard the
possibility that Pb loss and/or metamorphic recrystallization
induced the observed age spread as a real concern. Accord-
ingly, for these samples we report the measured age range
and refrain from making precise interpretations regarding
crystallization age.
4.1. Quimsa Cruz Intrusive Complex
[33] The Quimsa Cruz intrusion is composed of two
units. The main phase Mina Argentina (Mina Caracoles)
body consists of granodiorite and monzogranite. A second
phase situated to the SW is the Mina Viloco porphyritic
monzogranite with K-feldspar megacrysts [Miller, 1988;
Miller and Harris, 1989; Kennan et al., 1995]. Seventeen
of 18 zircon
206
Pb/
238
U ages from the Mina Argentina
sample (QCG5) and 10 of 11 results from the Mina Viloco
sample (QCG7) define distributions of late Oligocene
crystallization ages that are comparable in analytical scatter
(±3.5% and ±2.2% 1s) to that expected from homogeneous
samples based upon the reproducibility of standard analyses
during the run (±2%). From these results we calculate
Table 1. Summary of U-Pb,
40
Ar/
39
Ar, and Fission Track Results
a
Sample Elevation, m
U-Pb
Zircon, Ma
40
Ar/
39
Ar Fission Track
Muscovite, Ma Biotite, Ma K-Feldspar, Ma Apatite, Ma Zircon, Ma
Northern (Amaguaya) Transect
Yani granite
RJG7 3042 42.9 ± 1.1 40.7 ± 0.6 43.7 ± 4.5 8.6 ± 0.7
RJG8 3343 267–217 52.4 ± 1.3 44.6 ± 0.8 10.5 ± 0.9
RJG9 3414 84.0 ± 2.1 39.4 ± 0.6 42.1 ± 3.1 10.0 ± 0.9
RJG1 3805 273–218 55.0 ± 2.0 44.9 ± 0.7 34.6 ± 1.4 16.2 ± 1.3
03RJG15 3976 109.4 ± 4.9 132.5 ± 3.0 47.6 ± 3.1 15.4 ± 1.5
RJG2 4074 99.4 ± 3.8 38.3 ± 0.7 10.3 ± 0.9
RJG3 4335 47.5 ± 0.8 17.9 ± 1.8
Illampu granite
03RJG16 4612 251–226 161.4 ± 3.9 55.2 ± 2.1 14.3 ± 1.2
RJG14 4421 148.3 ± 4.2 43.9 ± 0.8
RJG4 4714 191.8 ± 3.8 74.5 ± 2.1 10.4 ± 1.7
RJG5 5020 149.2 ± 4.6 130.1 ± 4.7 11.2 ± 3.7
RJG6 5363 192.9 ± 3.9 107.2 ± 3.8 18.2 ± 1.1
Central (Rio Zongo) Transect
Huayna Potosı´ granite
RZG19 5400 217.6 ± 3.5 16.3 ± 1.0 151.0 ± 12.0
RZG17 4360 241 – 218
Zongo granite
RZG11 2840 263 – 227
Southern (Quimsa Cruz) Transect
Quimsa Cruz granite (Mina Argentina)
QCG8 3559 23.3 ± 0.4
QCG7 3900 25.65 ± 0.41 23.4 ± 0.4 6.8 ± 0.6
QCG6 4438 23.9 ± 0.8 5.8 ± 0.5
Quimsa Cruz granite (Mina Viloco)
QCG5 4830 26.02 ± 0.41 23.9 ± 0.4 15.5 ± 1.3
QCG4 5108 24.6 ± 0.4 16.2 ± 2.4
Eastern Altiplano Basin
Aranjuez Formation tuffs
03Taa01 25.48 ± 0.40
03Taa02 26.69 ± 0.54
03Taa03 25.08 ± 0.54
03Taa04 27.40 ± 0.54
03Taa05 25.23 ± 0.42
03Taa06 33.13 ± 0.80
Salla Formation tuffs
03Tsal4 28.00 ± 0.68
03Tsal6 24.59 ± 0.39
a
Errors reported at the 1slevel. Most U-Pb values represent observed age range of noninherited zircon grains (see text); U-Pb ages for QCG7 and QCG5
are based on the weighted mean of noninherited zircon grains. The
40
Ar/
39
Ar ages are weighted mean ages. Fission track ages are pooled ages. See the
auxiliary material for complete analytical results.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
9of22
TC6007

weighted mean ages of 26.2 ± 0.2 Ma (QCG5; Figure 5e)
and 25.4 ± 0.2 Ma (QCG7; Figure 5f), respectively. Our
results conflict with less precise data that suggested Eocene
cooling for this pluton [Kennan et al., 1995; Lamb et al.,
1997]. Moreover, we did not confirm a proposed older
Permo-Triassic heritage [Miller, 1988; Miller and Harris,
1989; Lamb and Hoke, 1997]. The Mina Argentina sample
yielded a discordant late Paleozoic grain while the Mina
Viloco sample produced one discordant zircon with a
207
Pb/
206
Pb age of 2656 ± 6 Ma. This Late Archean age
matches older provinces of the Amazon craton to the east
[e.g., Santos et al., 2000] and could conceivably imply
inheritance from cratonic rocks underthrust beneath the
central Andean plateau [e.g., Beck and Zandt, 2002].
4.2. Illampu Granite
[34] Eight of 10 zircons analyzed from the Illampu
granite (03RJG16) yield Late Permian to Late Triassic
206
Pb/
238
U ages (251–226 Ma). Uranium concentrations
are in excess of 1000 ppm with the highest U grain
(3500 ppm) yielding a Th/U value of 0.004 that is typical
of metamorphic recrystallization. We consider it possible
that the Illampu granite was emplaced as early as 250 Ma
and that the younger ages reflect Pb loss and metamorphic
recrystallization. Two Silurian U-Pb ages (430 Ma) were
also measured (Figure 5a) from low U grains (<500 ppm)
with Th/U values typical of magmatic conditions (>0.1).
These U-Pb results are similar in age to reset metamorphic
crystalline basement in the Western Cordillera [Wo¨ rner et
al., 2000] and probably represent assimilated wall rock.
4.3. Yani Granite
[35] Results from the Yani granite (RJG1) are similar to
those from the Illampu body in that 10 of 18 zircons define
a range of
206
Pb/
238
U ages between 273 and 218 Ma
(Figure 5b). Similar to the Illampu Granite, some grains
with high U and low Th/U content are present. However,
relatively low U grains (<500 ppm) with magmatic Th/U
(>0.1) values also occur in the youngest portion of the age
distribution, at 230 –218 Ma. Hence the implications of the
results for determining the emplacement age of the Yani
granite are ambiguous and more detailed work is required.
Nevertheless, it seems clear that the granite was likely
intruded during Permo-Triassic time. The remaining
8 zircons yield variably discordant U-Pb ages that appear
to reflect assimilation of basement of early Paleozoic and
Late Proterozoic age, similar to the Altiplano and Western
Cordillera [Lehmann, 1978; Wo¨rner et al., 2000].
4.4. Huayna Potosı´ Granite
[36] Ten of 12 zircon
206
Pb/
238
U ages measured from the
Huayna Potosı´ granite (RZG17) define a range from 241 to
218 Ma (Figure 5c). Similar to the previous sample, some
high U, low-Th/U analyses were obtained along with lower
U analyses characterized by magmatic Th/U values. More
detailed work is required to more rigorously assess the
emplacement age of the Huayna Potosı´ granite. The two
older analyses are slightly discordant and appear to repre-
sent assimilated early Paleozoic basement.
4.5. Zongo Granite
[37] Eleven of 15 zircons measured from sample RZG11
from the Kuticucho phase of the Zongo granite [McBride et al.,
1987] define a range of
206
Pb/
238
U ages from 263 to 227 Ma
(Figure 5d). The combination of high U concentrations
(typically 1000 ppm) and low Th/U values (generally
<0.03) indicate a high probability that Pb loss and/or
metamorphic recrystallization significantly affect the results.
The remaining 4 analyses are variably discordant and appear
to represent assimilated early Paleozoic and Late Proterozoic
basement rocks.
5. The
40
Ar/
39
Ar Thermochronology
[38] The
40
Ar/
39
Ar step-heating analyses were carried out
on 18 granite and 8 tuff samples (Table 1). A more detailed
accounting of the methods employed and results obtained is
provided in Table S2 (see auxiliary material). Uncertainties
are reported at the 1slevel and include analytical errors and
uncertainties in J factors.
[39] The new
40
Ar/
39
Ar data help constrain cooling
histories for the Cordillera Real granitoids and eruption
ages for volcanic tuffs within the Cenozoic succession of
the eastern Altiplano. Previous investigations of granites in
the Cordillera Real and along strike in southern Peru
identified complex thermal histories believed to have
resulted from Eocene thrust-induced heating and subsequent
cooling [Evernden et al., 1977; McBride et al., 1983, 1987;
Farrar et al., 1988; Heinrich, 1988; Kontak et al., 1990;
Sandeman et al., 1995]. This interpreted tectonothermal
episode of transient heating was considered to have pro-
duced anomalous relationships (i.e., biotites yielding older
ages than corresponding muscovites [McBride et al., 1987;
Farrar et al., 1988; Kontak et al., 1990]). New
40
Ar/
39
Ar
results reveal a more straightforward pattern of mica and K-
feldspar
40
Ar/
39
Ar ages that can be interpreted as the
product of exhumation-related cooling. Finally, while pre-
vious studies have dated volcanic horizons in the eastern
Altiplano basin [e.g., MacFadden et al., 1985; Kay et al.,
1998], the age of the oldest synorogenic fill remained
poorly constrained. New
40
Ar/
39
Ar biotite ages presented
here indicate that the sediments were deposited several
million years earlier than previously indicated.
5.1. Northern Transect
[40] The 12 granite samples along the northern transect
yield
40
Ar/
39
Ar total gas ages that become progressively
younger with decreasing elevation. This overall relationship
is similar to that determined in previous studies [e.g.,
Farrar et al., 1988; Kontak et al., 1990]. From SW to NE
(5363 to 3042 m), muscovite ages decrease from 193 ± 4 to
43 ± 1 Ma. Similarly, biotite ages decrease from 133 ± 3 to
39 ± 1 Ma (Figure 6 and Table 1). Unfortunately, features
exhibited by the age spectra indicate variable contamination
with excess
40
Ar (
40
Ar
E
)[Roddick et al., 1980; Dallmeyer
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
10 of 22
TC6007

and Rivers, 1983; Zeitler and Fitz Gerald, 1986; Foster et
al., 1990; Baxter et al., 2002] (Figure 7). Results of
isothermal duplicate and quadruplicate experiments in the
initial heating steps for K-feldspar confirm retention of
excess argon in the lower gas release portions for K-
feldspars examined in this study. These materials yield
saddle-shaped age spectra that indicate
40
Ar
E
contamination
affects both low- and high-temperature gas release (Figure 8)
[e.g., Harrison et al., 1994].
[41] In general, biotite
40
Ar/
39
Ar age spectra exhibited
less evidence of
40
Ar
E
contamination at the lower elevation
localities in the NE (i.e., below 4400 m elevation). These
biotites yielded younger ages and flatter age spectra
(Figure 7a) than biotites collected from above 4400 m in
the SW (Figure 7b). Muscovite results exhibited similar
patterns. For example, the youngest NE muscovite (RJG7)
yields a flat age spectrum with little evidence of distur-
bance. In contrast, high-elevation muscovites from the SW
yield saddle-shaped spectra (Figure 7c).
[42] Although
40
Ar
E
contamination clearly limits our
ability to carry out detailed interpretation of the
40
Ar/
39
Ar
data, several considerations suggest that the overall extent
of contamination was not sufficient to preclude meaningful
interpretation of the overall nature of the results. For
example, none of the micas are so contaminated with
40
Ar
E
that
40
Ar/
39
Ar ages exceed corresponding U-Pb
crystallization ages. Also, coexisting muscovite and biotite
generally yield similar ages. In 10 of 11 cases, the musco-
vites were older than coexisting biotite, as expected from
their high intrinsic Ar retentivity. This relationship contrasts
with that reported for the Zongo-San Gaba´n region extend-
ing from the Cordillera Real 450 km to the NW into
southern Peru where biotites are apparently so contaminated
with
40
Ar
E
that their K-Ar ages commonly exceed those of
coexisting muscovite [McBride et al., 1987; Farrar et al.,
1988; Kontak et al., 1990]. Finally, the lowest elevation
sample along the transect yields highly correlated inverse
isochrons for muscovite and biotite consistent with nearly
atmospheric trapped argon (Table S2; see auxiliary material).
Hence we believe that at least the topographically lowest
samples yield geologically meaningful ages that constitute a
robust record of Cenozoic cooling within the study region.
For biotite, the younger ages (8 of 12 samples) fall between
38.3 ± 0.7 and 55.2 ± 2.1 Ma. The younger muscovite ages
(3 of 11 samples) range from 42.9 ± 1.1 to 55.0 ± 2.0 Ma.
As detailed below, the Eocene mica bulk closure ages can
be explained by acceleration of denudation rates in the
middle Eocene with continued high rates persisting into the
Oligocene.
[43] To further quantify Cenozoic cooling for the north-
ern transect, we also measured K-feldspar
40
Ar/
39
Ar age
spectra and interpreted the results using the multidiffusion
domain (MDD) model [Lovera et al., 1989, 1997, 2002;
Lee, 1995; see also Parsons et al., 1999]. Because the K-
feldspar age spectra are seriously affected by
40
Ar
E
,we
limited our modeling efforts to the least affected portion of
Figure 6. Age-distance plot showing
40
Ar/
39
Ar musco-
vite, biotite, and K-feldspar results and apatite fission track
(AFT) data for 12 samples along the northern transect. At
this scale, the individual 1sage errors are contained within
the various geometric symbols.
Figure 7. The
40
Ar/
39
Ar age spectra for biotite and muscovite grains from granite samples of the
northern and southern transects. Note variable scale on vertical (age) axes. (a) Northern transect biotite
(<4400 m). (b) Northern transect biotite (>4400 m). (c) Northern transect muscovite. (d) Southern
transect biotite.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
11 of 22
TC6007

the age spectra (i.e., intermediate release steps [see Zeitler
and Fitz Gerald, 1986; Lovera et al., 2002]). MDD mod-
eling of four K-feldspars from the Yani granite at lower
elevations of the northern transect (Figure 8) constrain the
mid-Cenozoic thermal histories and help link the higher
temperature
40
Ar/
39
Ar mica results to the apatite fission
track results discussed below. The MDD model results
(Figure 8) demonstrate 8 –12°C/Ma cooling between about
38 and 26 Ma with average cooling rates generally increas-
ing with decreasing elevation from SW to NE.
5.2. Southern Transect
[44] Biotite
40
Ar/
39
Ar analyses from five Quimsa Cruz
samples along a vertical interval of 1550 m (Figures 3
and 7d) yield late Oligocene–earliest Miocene cooling
ages ranging from 24.6 ± 0.4 to 23.3 ± 0.4 Ma (Table 1).
The samples produce relatively flat age spectra (Figure 7d)
and become younger with decreasing elevation to the NE. The
fact that the biotite
40
Ar/
39
Ar results are slightly younger than
the 26 Ma U-Pb zircon crystallization age measured from
the Quimsa Cruz granite is consistent with a depth of
emplacement of >4 km [see also Miller, 1988; Lamb and
Hoke, 1997].
5.3. Depositional Age Constraints for the Altiplano
Basin
[45] Biotite
40
Ar/
39
Ar results are reported for 8 volcanic
tuffs in the Cenozoic succession of the eastern Altiplano
(Figure 9 and Table 1). A discontinuous belt of exposed
basin fill occurs between 16 and 17.5°S, including the Pen˜as
and Aranjuez formations in the north and the Luribay and
Salla formations in the south (Figure 1b) [Martinez, 1980;
Geobol, 1993, 1995; Sua´rez and Diaz, 1996].
[46] The deposits are 500 –1500 m thick, but only the
upper levels of the southern exposures have been studied in
detail, where previous fission track,
40
Ar/
39
Ar, and magne-
tostratigraphic studies have revealed Oligocene deposition
of the Salla Formation from 29.4 to 25.5 Ma [MacFadden et
al., 1985; McRae, 1990; Kay et al., 1998]. New results for
tuffs in the lower to intermediate levels include six tuffs of
the Aranjuez Formation ranging from 33.1 ± 0.8 to 25.2 ±
0.4 Ma (Table 1). Two additional tuffs from the Salla
Formation yield ages of 28.0 ± 0.7 and 24.6 ± 0.4 Ma.
These results demonstrate that sediment accumulation had
already commenced in the eastern Altiplano by the Eocene-
Oligocene boundary, earlier than commonly envisioned
[e.g., Sempere et al., 1990].
6. Fission Track Thermochronology
[47] Fission track analysis was employed to further
evaluate the low-temperature (<120°C) cooling history of
the Cordillera Real. Apatite fission track (AFT) data help
constrain the time at which rocks were exhumed above
4 km depth (110–120°C), assuming a steady geothermal
gradient. Moreover, their track length distributions can be
modeled to constrain low-temperature (60 –120°C) ther-
mal histories [Laslett et al., 1987; Lutz and Omar, 1991;
Willett, 1997; Ketcham et al., 2000]. The new results build
upon those of previous studies, notably Benjamin et al.
[1987], who presented AFT and zircon fission track (ZFT)
results along the central (Rio Zongo) transect. Additional
fission track data have also been presented within the
context of several other studies [Crough, 1983; Heinrich,
1988; Safran, 1998]. The previous results have given rise to
conflicting interpretations regarding the onset and rates of
cooling in the region [Benjamin, 1986; Benjamin et al.,
1987; Masek et al., 1994; Lamb and Hoke, 1997; Safran,
1998; Moore and England, 2001; Anders et al., 2002].
[48] Seventeen fission track analyses were performed on
16 granite samples, including 11 AFT results for the
northern transect, 4 AFT results for the southern transect,
Figure 8. (left) The
40
Ar/
39
Ar age spectra and (right)
multidomain diffusion (MDD) modeling results for K-
feldspar grains from Yani granite samples of the northern
transect. (a) Sample 03RJG15. (b) Sample RJG1.
(c) Sample RJG9. (d) Sample RJG7.
Figure 9. The
40
Ar/
39
Ar age spectra for biotite grains from
volcanic tuffs of the eastern Altiplano basin. (a) Results for
six tuff samples from the Aranjuez Formation. (b) Results
for two tuff samples from the Salla Formation.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
12 of 22
TC6007

and AFT and ZFT results for a single sample from the
central transect. AFT ages range from 18.2 to 5.8 Ma, with
significant age overlap among the transects (Figure 6 and
Table 1). Ages are reported as pooled values with 1serrors.
Complete results are reported in Table S3 (see auxiliary
material).
[49] Modeling of the new AFT data yields results con-
sistent with those obtained from higher-temperature
40
Ar/
39
Ar thermal history data (i.e., acceleration of denuda-
tion rates in the middle Eocene with continued high rates
persisting into the Oligocene). Similar to
40
Ar/
39
Ar data, the
fission track results show progressively younger ages from
SW to NE, from high to low elevation (Figure 6). For the
northern transect, the oldest and youngest AFT ages of 18.2
and 8.6 Ma are from the highest and lowest (5363 and
3042 m) samples, respectively. AFT ages for the southern
transect generally correlate with elevation, from 16.2 to
5.8 Ma between 5108 and 3900 m. The ZFT age of 151 Ma
for the central transect provides a critical data point for the
Huayna Potosı´ granite at 5400 m, clarifying previous
interpretations of Benjamin et al. [1987]. A correlation
between AFT age and elevation throughout the Cordillera
Real suggests a simple record of exhumation-induced cool-
ing. However, a few samples depart slightly from this trend,
suggestive of locally more complex histories possibly
related to a variable thermal structure at shallow depth, as
dictated by topographic relief.
6.1. Northern Transect
[50] AFT analyses were performed on 11 of 12 samples
from the northern transect. The AFT results (Table 1) show
Miocene cooling, with ages ranging from 18.2 ± 1.1 Ma for
the highest SW sample to 8.6 ± 0.7 Ma for the lowest NE
sample (Figure 6). Mean track lengths vary from 11.83 ±
0.28 to 14.06 ± 0.20 mm, with standard deviations of
generally 1.5–2.7 mm. AFT ages are generally older at
higher elevation although a few exceptions to this trend are
interspersed along the transect (Table 1). The younger
samples generally have longer mean track lengths and
overall more restricted length distributions (Table 1) indi-
cating more rapid cooling [Green, 1986; Fitzgerald, 1994].
6.2. Southern Transect
[51] AFT analyses of the four Quimsa Cruz samples yield
ages ranging from 16.2 ± 2.4 to 5.8 ± 0.5 Ma (Table 1). The
younger samples are located at lower elevations in the NE
while the older ages are generally from the highest part of
the transect in the SW. Relative to the northern transect, the
mean track lengths tend to be slightly longer, ranging from
13.19 ± 0.24 to 13.75 ± 0.26 mm (Table 1). However,
average track length distributions are comparable to slightly
broader, with standard deviations of 2.19–2.66 mm, sug-
gesting potentially more complex thermal histories [e.g.,
Fitzgerald and Gleadow, 1988]. When considered together
with the
40
Ar/
39
Ar biotite results, the fission track data
clearly support late Oligocene through late Miocene slow
exhumation of the Quimsa Cruz granite. Although the data
set is limited, the younger AFT ages are consistent with a
phase of accelerated denudation beginning at 6 Ma.
6.3. Modeling of AFT Results
[52] Permissible AFT cooling histories were constrained
by the measured track length distributions and kinetic
parameters derived from etched track pit diameters (D
par
values) [Carleson et al., 1999; Donelick et al., 1999;
Ketcham et al., 1999, 2000]. Measured track lengths have
been corrected using a C axis projection [Donelick et al.,
1999] and each sample was modeled assuming a single
kinetic population, consistent with the observed narrow
range in D
par
values (1.51–1.64 mm). Confidence limits
of the cooling paths were produced using a Monte Carlo
search algorithm, yielding an ‘‘acceptable fit’’ envelope
(statistical probability value = 0.05), and a ‘‘good fit’’
envelope (statistical probability value = 0.5) [Ketcham et
al., 2000].
[53] AFT modeling for the northern transect reveals nearly
linear, <120°C time-temperature paths indicative of uniform
rapid cooling from late Miocene onward (Figure 10a).
In contrast, older samples display variable cooling histories.
Several older samples (notably 03RJG16 and RJG3) sug-
gest slower cooling during the early to middle Miocene
followed by rapid late Miocene cooling (Figure 10b). The
remaining older samples (RJG6, 03RJG15, and RJG1)
exhibit approximately linear trends suggesting uniform
cooling from late Oligocene–early Miocene to Pliocene
time.
[54] Approximate cooling rates were estimated for mod-
eled ‘‘good fit’’ cooling envelopes between 110 and 60°C.
The younger, 8.6 – 11.9 Ma samples (Figure 10a) yield
cooling of 7–12°C/Myr between 10 and 6 Ma. The
older samples (Figure 10b) exhibit cooling of generally
<5°C/Myr between 23 and 8 Ma. For the southern
transect, AFT thermal models indicate cooling of <5–
11°C/Myr from 19 to 9 Ma, with more rapid late
Miocene cooling of <5–20°C/Myr. Overall, the AFT ages
and model results (Figure 10) indicate rapid cooling <120°C
during late Miocene to Pliocene time, with some samples
suggesting earlier slower cooling during early to middle
Miocene time. Moreover, the results are quite consistent
with the K-feldspar
40
Ar/
39
Ar results in that they indicate
continued rapid cooling during Eocene-Oligocene time.
7. Cenozoic Exhumation of the Cordillera Real
7.1. Integrated Thermal Histories
[55] Thermochronologic results help constrain the cool-
ing history of the Cordillera Real. Composite thermal
histories were constructed using
40
Ar/
39
Ar results and
relevant closure temperatures for muscovite (400 –345°C)
and biotite (325–280°C) [McDougall and Harrison, 1999],
MDD modeling of K-feldspar
40
Ar/
39
Ar results (250–
150°C) [Lovera et al., 1989], and modeling of apatite
fission track length distributions (120 –60°C) [Ketcham et
al., 2000].
[56] Four samples from the northern transect (03RJG15,
RJG1, RJG9, and RJG7) were targeted on the basis of (1)
lower excess argon retention, suggesting more reliable,
longer duration MDD histories, (2) lower uncertainties for
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
13 of 22
TC6007

AFT modeling results, and (3) the internal consistency
afforded by samples from the same rock unit, the Yani
granite. The MDD and AFT cooling histories are plotted in
time-temperature space, along with the
40
Ar/
39
Ar mica ages
(Figure 11). High-temperature portions of the AFT models
were constrained by requiring the composite thermal histo-
ries to fit the lowest, well-defined temperature envelope of
the MDD models.
[57] Higher elevation samples in the SW (03RJG15 and
RJG1) exhibit protracted histories that extend back to
Cretaceous–early Cenozoic time. Their cooling paths show
pronounced slope breaks in time-temperature space indica-
tive of an onset of rapid cooling during the middle to late
Eocene (40–35 Ma; Figures 11a– 11b). The highest sample
shows uniform cooling since 37 Ma at 6°C/Myr while the
other suggests a slight reduction in cooling rate at 26 Ma,
from 10 to 5°C/Myr.
[58] The topographically lower NE samples (RJG9 and
RJG7) record more abbreviated histories that commenced at
45–40 Ma. Steeper segments of the cooling paths indicate
two phases of rapid cooling: an initial middle Eocene –
Oligocene phase that was underway by 40 Ma and
persisted until 26 Ma and a second phase from late
Miocene onward (Figures 11c– 11d). The cooling paths
reveal initial rapid cooling between about 43 and 26 Ma
at 11°C/Myr and later rapid cooling at 9–16°C/Myr from
11 Ma onward. A period of nearly isothermal conditions
appears to separate the two phases of rapid cooling. This
episode of diminished cooling is further supported by
AFT model results for additional samples, notably RJG3
(Figure 10b), which indicates <1.0 – 2.0°C/Myr cooling
from 22 to 8 Ma.
[59] In summary, the integrated cooling histories uni-
formly identify a middle Eocene inception of rapid cooling.
Figure 10. Apatite fission track (AFT) (left) thermal modeling results and (right) track length data for
(a) younger samples and (b) older samples. Each AFT model shows the calculated time-temperature path,
pooled age, and modeled age (in parentheses). Each track length plot shows the track length distribution
(histogram), modeled distribution (curve), observed mean track length, and modeled mean track length
(in parentheses). All reported track lengths represent corrected values using a C axis projection [Donelick
et al., 1999].
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
14 of 22
TC6007

This phase of rapid cooling (11°C/Myr) persisted from
45–40 to 26 Ma. Lower-elevation samples further con-
strain a period of reduced cooling from 26 to 11 Ma
(0–2°C/Myr) followed by rapid cooling (9–16°C/Myr)
from 11 Ma onward.
7.2. Age-Elevation Relationships
[60] Examination of the variation of mica
40
Ar/
39
Ar ages
as a function of elevation further constrains the higher
temperature cooling history of the Cordillera Real. Provided
an increase in cooling rate, samples will define a broad age
range at high elevation that narrows with decreasing eleva-
tion. These age patterns are produced by the intersection of
present-day topography and paleodepth as manifested by
the exhumed partial retention zone for Ar in the micas.
When not obscured by
40
Ar
E
contamination, the onset of
rapid cooling will be marked as an abrupt inflection in a plot
of age vs. elevation that defines the lower limit of the partial
retention zone (Figure 12a) [Fitzgerald et al., 1995].
[61] This predicted relationship is observed within the
Cordillera Real in spite of complications related to
40
Ar
E
contamination at high elevation. The
40
Ar/
39
Ar age differ-
ences for 11 muscovite-biotite pairs along the northern
transect become systematically smaller (from 86 to
2 Myr) from high to low elevations. This relationship
suggests the low-elevation micas were sufficiently deep that
they were completely open with respect to Ar retention and
closed only as a result of initial rapid cooling during the
middle Eocene (Figure 6). We interpret the transition from
distributed biotite ages (130.1 –55.2 Ma) to clustered ages
(47.5–38.3 Ma) below 4500 m elevation as indicating the
onset of rapid cooling at 45 –40 Ma (Figure 12b). Assuming
a steady 30°C/km geothermal gradient [Benjamin et al.,
1987; Henry and Pollack, 1988], the biotite results indicate
that the average cooling rate increased at 45 –40 Ma from
<5 to 10°C/Myr.
40
Ar/
39
Ar weighted mean ages for biotite,
muscovite, and K-feldspar from the lowest sample (RJG7)
are nearly identical at 44–40 Ma (Figure 6 and Table 1),
further supporting rapid cooling at this time.
[62] On an age-elevation plot, the broad distribution of
muscovite cooling ages (Figure 12b) reflects relatively slow
cooling at <5°C/Myr prior to middle Eocene time. An
inflection in the trend of the data suggests that samples
above 3300 m resided in the muscovite partial retention
zone prior to 45 Ma. A similar inflection in the biotite data
suggests that an exhumed partial retention zone for biotite is
represented by samples above 4400 m (Figure 12b). These
data require that the samples resided in a narrow zone at 9–
11.5 km depth prior to 45 Ma, given the assumptions of
closure temperature and geothermal gradient (Figure 11a).
This estimate is in agreement with the reported 10–15 km
granite emplacement depth (3 – 5 kbar) [McBride et al.,
1983; Heinrich, 1988].
[63] A middle Eocene inception of rapid cooling is
consistent with ZFT data for the central transect [Benjamin
et al., 1987] (Figure 12c). Benjamin et al. [1987] speculated
that their oldest (176.9–62.8 Ma), highest elevation (5100 –
3900 m) samples cooled within the ZFT partial annealing
zone prior to more rapid cooling at 45 Ma. A new ZFT
analysis of a sample from 5400 m elevation (Table 1),
300 m above their highest sample, supports this interpre-
tation. The data define a trend from widely distributed ZFT
ages (roughly 177–63 Ma) at higher elevations to a narrow
distribution of ages (mostly 45 –32 Ma) below 3900 m
(Figure 12c). In addition, the younger ZFT ages of 36.0–
31.9 Ma [Benjamin et al., 1987] are similar to the youngest
K-feldspar
40
Ar/
39
Ar age of 34.6 ± 1.4 Ma for the northern
transect (Table 1). These comparable ages, despite the
disparity in closure temperatures, suggest that rapid cooling
continued during late Eocene to Oligocene time. Moreover,
the ZFT and
40
Ar/
39
Ar mica data support the MDD and
composite modeling results (Figures 8 and 11) indicating
initial rapid cooling at 45–40 Ma.
[64] On an age-elevation plot, AFT data for the northern
transect define a steep, nearly linear trend suggesting
uniform rapid cooling was underway by about 10 Ma
(Figure 12b). The absence of more broadly distributed data
at higher elevations suggests that the partial annealing zone
has been removed by erosion. Modest AFT age variations
reveal that some anomalous older samples locally contradict
the trend of increasing age at higher elevations (Figure 12b),
as evidenced in a NE-SW profile (Figure 6). Several
Figure 11. Composite thermal modeling results (com-
bined
40
Ar/
39
Ar MDD modeling and AFT thermal model-
ing) for four granite samples from the northern transect.
(a) Sample 03RJG15. (b) Sample RJG1. (c) Sample RJG9.
(d) Sample RJG7.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
15 of 22
TC6007

mechanisms may explain older samples interspersed with
younger samples. First, postcooling displacement along
faults could account for the age contradictions. However,
the older ages generally occur in hanging wall settings,
inconsistent with regional thrust faulting [Pareja et al.,
1978; Martinez, 1980; McQuarrie, 2002]. Second, because
the 110– 120°C isotherm is sensitive to topography at small
wavelengths [e.g., Stu¨we et al., 1994], particularly in the
high-relief regions such as the Cordillera Real [Safran,
1998], topographic effects on the former geothermal gradi-
ent may produce the necessary age variations. Third, the age
discrepancies may be explained by compositional differ-
ences in apatite, as reflected by D
par
(etch pit diameter)
values, which affect the closure temperature of individual
samples [Carleson et al., 1999]. Despite the narrow range of
D
par
values (1.51–1.64 mm), there is a positive relationship
between higher D
par
values (>1.60 mm) and older ages
(Figure 10). We tentatively attribute the anomalous older
AFT ages to the combined effects of topography and
compositional variations.
[65] For the southern transect, the cooling history is
constrained by limited
40
Ar/
39
Ar biotite and AFT results.
On an age-elevation plot, the biotite ages reveal a steep,
nearly linear trend at about 25–23 Ma (Figure 12d).
Because U-Pb zircon ages reveal a 26 Ma crystallization
age for the Quimsa Cruz granite, the biotite suggest that
modest rates of erosional denudation from >4 km depth
accompanied pluton emplacement. AFT age data from four
Quimsa Cruz samples define a stair step curve with an
inflection suggestive of an increase in cooling rate at 6Ma
(Figure 12d).
7.3. Eastern Altiplano Sedimentation
[66] The thermal record of exhumation is compatible with
the sedimentation record of the eastern Altiplano.
40
Ar/
39
Ar
ages of 8 tuffs (Figure 9 and Table 1) within the Cenozoic
succession help define the age of basin fill at 16 – 17.5°S,
including the Pen˜as, Aranjuez, Luribay, and Salla forma-
tions (Figure 1b) [Martinez, 1980; MacFadden et al., 1985;
McRae, 1990; Sempere et al., 1990; Geobol, 1993, 1995;
Sua´rez and Diaz, 1996; Lamb and Hoke, 1997; Kay et al.,
1998; McQuarrie and DeCelles, 2001]. The age data reveal
deposition spanning the Oligocene, from 33.1 to 24.6 Ma.
[67] Provenance data (Figure 13) for these deposits reveal
the composition of source areas and the sediment transport
pathways. Sandstone petrographic analyses consisted of
point counts (450 grains per thin section) of 11 samples
Figure 12. (a) Schematic age-elevation plot of idealized
distribution of ages associated with an exhumed partial
retention/annealing zone (PRZ/PAZ) [after Fitzgerald et al.,
1995]. Thermochronometer sample age-elevation plots of
40
Ar/
39
Ar muscovite, biotite, and K-feldspar weighted mean
ages, apatite fission track (AFT) pooled ages, and zircon
fission track (ZFT) pooled ages for the (b) northern transect,
(c) central transect, and (d) southern transect. The shaded
lines represent the approximate age-elevation trends for the
various thermochronometers.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
16 of 22
TC6007

from a 200 m stratigraphic interval in the Luribay For-
mation to middle Salla Formation. A modified Gazzi-
Dickinson method of point counting was employed
[Ingersoll et al., 1984] and complete results are presented
in Table S4 (see auxiliary material). Recalculated detrital
modes are plotted in ternary diagrams depicting proportions
of Q-F-L (quartz-feldspar-lithic fragments; Figure 13a) and
Lv-Lm-Ls (lithic volcanic-metamorphic-sedimentary frag-
ments; Figure 13b). Both diagrams show upsection varia-
tions in detrital composition. The Q-F-L plot displays a
minor decrease followed by a dramatic increase in lithic frag-
ments (Figure 13a). The Lv-Lm-Ls plot depicts an upsection
shift from sedimentary to metamorphic fragments followed
by a huge increase in the proportion of volcanic fragments
(Figure 13b). Quartz, sedimentary, and metamorphic frag-
ments are consistent with derivation from Paleozoic clastic
sedimentary and metasedimentary rocks of the Eastern
Cordillera. The Luribay-Salla sandstones generally contain
a greater proportion of lithic fragments, particularly volcanic
fragments, relative to previous Altiplano studies [Horton
et al., 2002], possibly related to closer proximity to sedi-
ment sources. In any case, the increased proportion of
volcanic detritus is considered a record of increasing vol-
canism in the Cordillera Real, likely associated with em-
placement of the Quimsa Cruz granite.
[68] Over 330 paleocurrent indicators were measured,
including trough axes and trough limbs. Paleocurrent data
for the Aranjuez and Luribay-Salla localities indicate trans-
port to the SSW and NW, respectively (Figures 13c and 13d).
Similar to the compositional data, these dispersal patterns are
compatible with derivation from the neighboring Cordillera
Real.
[69] Structural relationships shed further light on Ceno-
zoic basin evolution. Discontinuous exposures of the Pen˜as,
Aranjuez, and Luribay-Salla formations aligned parallel to
the Cordillera Real (Figure 1b) share similar ages, prove-
nance, facies, and unconformable basal contacts on steeply
dipping Silurian-Devonian strata. Growth strata within the
Luribay Formation and lower levels of the 29.4– 25.5 Ma
Salla Formation indicate syndepositional activity along SW
directed fold-thrust structures prior to 28 Ma [Gillis et
al., 2004]. Growth geometries are not observed in the
subhorizontal strata of the upper succession, indicating late
Oligocene cessation of upper crustal deformation in the
region.
[70] The compositional and paleocurrent analyses dem-
onstrate that the Cordillera Real was an active sediment
source during early evolution of the eastern Altiplano basin.
40
Ar/
39
Ar age data and structural relationships indicate that
deposition was occurring from about 33 to 25 Ma, largely
coeval with SW directed thrusting, in agreement with rapid
Oligocene exhumation identified by thermochronologic
data. These relationships indicate that crustal shortening,
unroofing and rapid cooling were synchronous along the
eastern margin of the central Andean plateau, suggesting
that crustal shortening was a principal cause of exhumation
in the early orogenic history.
8. Discussion
[71] Thermochronologic data help determine the history
of exhumation-related cooling in the Cordillera Real, but a
significant challenge lies in interpretation of the deforma-
tion history associated with this thermal record. In consid-
ering alternative tectonics models for the eastern flank of the
central Andean plateau, several observations must be
addressed.
8.1. Kinematics
[72] Structural data and crosscutting relationships restrict
the magnitude and timing of shortening on upper crustal
structures. Most notable is the limited displacement (<2 –
5 km) accommodated by most thrusts, suggesting 20 km of
minimum NE-SW shortening (35% net strain) across the
Cordillera Real (Figure 4). No evidence exists for near-
surface large-magnitude thrusts. Nevertheless, the range
exposes some of the deepest structural levels in Bolivia,
implying a larger structure at depth near the Eastern
Cordillera–Altiplano boundary [e.g., Dorbath et al., 1993;
McQuarrie, 2002].
[73] In terms of kinematics, upper crustal deformation
postdated Permo-Triassic magmatism and early Paleocene
deposition of the El Molino Formation (Figure 1b) [Lamb
and Hoke,1997;DeCelles and Horton, 2003; Horton,
2005]. Tighter age constraints are provided by growth strata
indicating SW directed thrusting synchronous with Oligo-
cene, and possibly late Eocene, sedimentation [Gillis et al.,
Figure 13. Provenance data for basin fill of the eastern
Altiplano basin. (a) Quartz-feldspar-lithic fragments (Q-F-
L) and (b) lithic volcanic-metamorphic-sedimentary frag-
ments (Lv-Lm-Ls) ternary diagrams showing data and
upsection trends (dashed arrows) for the Luribay-Salla
succession. Paleocurrent rose diagrams and vector means
for the (c) Aranjuez Formation and (d) Luribay-Salla
succession.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
17 of 22
TC6007

2004].
40
Ar/
39
Ar tuff ages (Figure 9 and Table 1) and
sediment provenance information (Figure 13) suggest that
deformation and surface uplift were in progress along the
eastern flank of the central Andean plateau between 33 and
25 Ma. For younger age constraints, crosscutting relation-
ships and U-Pb ages (Figures 3 and 5 and Table 1) show
that fold-thrust deformation predated 26 Ma emplacement
of the Quimsa Cruz granite [Evernden et al., 1977; Lamb
and Hoke, 1997]. Although it is widely accepted that upper
crustal shortening in the Eastern Cordillera had ceased by
10 Ma, on the basis of the extensive San Juan del Oro
surface [Gubbels et al., 1993; Kennan et al., 1997], inter-
montane basins throughout the Eastern Cordillera suggest
that major upper crustal shortening was complete by latest
Oligocene to earliest Miocene time [Horton, 2005]. With
the emergence of improved timing constraints for the
Cordillera Real and broader Eastern Cordillera of Bolivia,
it is clear that Paleogene deformation is integral to the
evolution of the central Andean plateau [e.g., Horton et al.,
2001; DeCelles and Horton, 2003; Horton, 2005; McQuarrie
et al., 2005a], in contrast to many studies that consider
shortening to be younger than 25 Ma [Isacks, 1988;
Gubbels et al., 1993; Allmendinger et al., 1997; Kley and
Monaldi, 1998; Gregory-Wodzicki, 2000]. The observed map
relationships suggest limited upper crustal deformation after
25 Ma, but do not rule out the possibility of Neogene
passive transport of upper crustal material above deeper
structures.
8.2. Cooling History
[74] In addition to structural constraints, a viable model
must also explain thermochronologic results.
40
Ar/
39
Ar
mica ages and multidomain diffusion (MDD) modeling of
K-feldspar results demonstrate a middle Eocene initiation of
rapid cooling at 45–40 Ma, consistent with the previous
zircon fission track (ZFT) results of Benjamin et al. [1987],
as supplemented by a single new ZFT analysis (Figure 12c).
Thermal modeling of
40
Ar/
39
Ar and AFT results suggest
that initial rapid cooling continued through the Oligocene to
26 Ma, followed by a period of slower cooling, then
renewed rapid cooling commencing at 11 Ma (Figure 11).
The
40
Ar/
39
Ar data display a pronounced and systematic
decrease in muscovite and biotite ages from SW to NE
(Figure 6), consistent with a northeastward progression of
exhumation.
40
Ar/
39
Ar data suggest that all but the lowest-
elevation sample cooled within the muscovite partial reten-
tion zone, and the 4 highest samples cooled within the
biotite partial retention zone, requiring residence at 280 –
345°C prior to 45 Ma (Figure 12b). Given the assump-
tions of closure temperature and steady geothermal gradient,
an estimated 9–11.5 km of total exhumation has occurred
since 45–40 Ma, with 3.5 km of that since 11 Ma.
These values are consistent with inferred 10–15 km
granite emplacement depths [McBride et al., 1983; Heinrich,
1988; Miller, 1988]) and further suggest that the middle
Eocene–late Oligocene phase of rapid cooling accounts for
most of the exhumation.
8.3. Tectonic Exhumation Models
[75] Several kinematic scenarios are capable of linking
the histories of deformation and early exhumation along the
eastern flank of the central Andean plateau. One possibility
involves large-scale, southwestward tilting of the Cordillera
Real in the backlimb of a belt of NE directed (10–30°SW
dipping), crustal-scale thrusts (Figure 14a) [e.g., Heinrich,
1988; Sheffels, 1990; Baby et al., 1997]. This tilting would
produce the observed systematic decrease in cooling ages
from SW to NE, and satisfy the observed low magnitudes of
Figure 14. Alternative schematic models for the tectonic evolution of the Cordillera Real, depicting the
Eocene-Oligocene history of the 11 samples (circles) of the northern transect at three stages: pre-44 Ma;
44 Ma; and 25 Ma. (a) Large-scale, down-to-SW tilting of the Cordillera Real due to motion along NE
directed thrust structures. (b) SW directed thrusting linked to an in-sequence duplex system defined by
four thrust sheets (1–4). (c) SW directed transport of a thick thrust sheet over a NE dipping ramp during
tectonic wedging. Hatchures represent zone of possible duplexing.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
18 of 22
TC6007

slip along mapped faults. However, restoration of the
sample transect to its predeformational state, by untilting
the samples 10–30°, would require the samples to have
been originally arrayed over a 6 –13 km vertical extent
(Figure 14a). Such a configuration is at odds with the paired
muscovite-biotite
40
Ar/
39
Ar ages requiring that most sam-
ples originally resided within a narrow temperature window
at 9–11.5 km depth.
[76] A second possibility is that mapped SW directed
thrusts in the Altiplano at the SW front of the central
Andean backthrust belt (Figure 1b) represent a system of
high-magnitude thrusts that merge at depth to the NE into a
regional de´collement horizon (Figure 14b). If this large-
magnitude structural system involved a duplex in Ordovi-
cian and older rocks [e.g., McQuarrie et al., 2005b], it could
produce substantial exhumation. In this model, rocks orig-
inally at 5–11 km depth are carried up a system of ramps at
different stages during in-sequence thrusting from NE to
SW. Although this scenario satisfies the estimated total
exhumation for some samples, it is not consistent with the
northeastward decrease in
40
Ar/
39
Ar cooling ages and the
requirement that all samples originated at 9 –11.5 km depth.
[77] A third model similarly involves transport of a thick
thrust sheet over a large NE dipping ramp (Figure 14c). This
geometry is consistent with limited slip along most faults,
but invokes a single large-magnitude, SW directed structure
that has not been identified. Nevertheless, several candi-
dates exist, including a seismically imaged crustal-scale
boundary between the Altiplano and Eastern Cordillera
[Dorbath et al., 1993] that may correspond to the Cordillera
Real Fault Zone [Martinez, 1980], a poorly defined struc-
ture apparently marking the SW margin of the granite belt
(Figure 1b). Because of the lack of cutoffs, it is conceivable
that a structure along the SW flank of the granite belt
accommodated significant displacement. Alternatively, a
large-magnitude structure may be covered by young basin
fill or may have fed slip into the Coniri and Huarina faults
(Figure 1b) at the SW front of the central Andean backthrust
belt [Martinez, 1980; Sempere et al., 1990]. These faults
place Devonian on Cenozoic rocks, but their total displace-
ment is uncertain.
[78] We tentatively favor the third model (Figure 14c),
which is congruent with the structural and thermochrono-
logic records. In considering these models, it is important to
note that the proposed large-magnitude structures may link at
depth with a deeper, NE directed de´collement (Figure 14c),
creating a crustal wedge geometry [e.g., Roeder, 1988;
Sempere et al., 1990; McQuarrie and DeCelles, 2001;
McQuarrie, 2002; McQuarrie et al., 2005a, 2005b].
8.4. Postshortening Exhumation
[79] On the basis of crosscutting relationships, most
upper crustal shortening in the Cordillera Real (this study)
and broader Eastern Cordillera of Bolivia [e.g., Horton,
2005] had ceased by 25 Ma. It is difficult to reconcile this
record with substantially increased cooling from late Mio-
cene onward, raising the possibility that processes other
than upper crustal shortening played a significant role in
youthful exhumation of the central Andean plateau. Poten-
tial mechanisms include (1) lower crustal thickening by
either thrust stacking [Sheffels, 1990; McQuarrie and
DeCelles, 2001; McQuarrie, 2002], ductile deformation
[Isacks, 1988], or underplating [Baby et al., 1997] coeval
with upper crustal shortening in the eastern foreland, (2)
lithospheric delamination beneath the Altiplano–Eastern
Cordillera boundary [Lamb and Hoke, 1997; Beck and
Zandt, 2002; Garzione et al., 2006], and (3) coupled ero-
sional denudation and rock uplift along the eastern plateau
margin [Masek et al., 1994; Horton, 1999; Montgomery
et al., 2001].
[80] Although thermochronologic and structural results
from this study are insufficient to distinguish between these
models, regional considerations suggest important differ-
ences between the northern and southern parts of the central
Andean plateau. Similar to our findings, AFT results across
the Eastern Cordillera 600 km to the south at 21°S
indicate predominantly Eocene–Oligocene cooling related
to crustal shortening [Ege et al., 2003; Ege, 2004]. These
data, however, lack signatures of the late Miocene and
younger exhumation recorded in the Cordillera Real at
15.5–17.5°S.
[81] Climate is similarly variable along strike, with mark-
edly higher precipitation, erosional denudation, and topo-
graphic relief north of 17.5°S contrasting with the arid,
low-relief region to the south [e.g., Masek et al., 1994;
Horton, 1999; Montgomery et al., 2001]. First-order con-
trols on precipitation include zonal atmospheric circulation
and orographic effects that focus precipitation along the
eastern flank of the Andes north of 17.5°S. This enhanced
precipitation has promoted localized rock uplift, high relief,
and headward retreat of the plateau margin [Masek et al.,
1994]. We propose that a pronounced north-south gradient
in precipitation and erosion may explain rapid late Miocene
and younger exhumation of the Cordillera Real in the
absence of upper crustal shortening. The reason for a late
Miocene inception of amplified exhumation is unknown,
but may relate to climate change or deeper crustal/litho-
spheric tectonics.
9. Conclusions
[82] Integration of structure, geochronology, thermochro-
nology, and sediment provenance demonstrates phases of
rapid exhumation-induced cooling along the eastern margin
of the central Andean plateau in Bolivia.
[83] 1. Mapping along three transects in the Cordillera
Real reveals principally SW directed thrusts of the central
Andean backthrust belt involving Paleozoic strata and
Mesozoic-Cenozoic granites. Individual faults exhibit low
magnitudes of displacement (<2–5 km). An estimated
minimum cumulative slip of 20 km, or 35% NE-SW
strain, is compatible with regional shortening estimates.
Nevertheless, exposure of deep levels along the eastern
flank of the central Andean plateau appears to require a
deeper structure that may be linked to either SW or NE
directed thrusts of the middle to lower crust.
[84] 2. Deformation timing is constrained by crosscutting
relationships among granites, fold-thrust structures, and
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
19 of 22
TC6007

basin fill. U-Pb analyses indicate Permo-Triassic and late
Oligocene (26 Ma) ages of granite crystallization. Permo-
Triassic granites and lower Paleocene strata predate defor-
mation, but the 26 Ma Quimsa Cruz granite postdates
most upper crustal shortening. Growth strata in the Luribay-
Salla succession further indicate SW directed thrusting
synchronous with pre-28 Ma sedimentation. Most fold-
thrust deformation is therefore considered to be of Eo-
cene-Oligocene age, in contrast to many previous studies
suggesting strictly Neogene age deformation in the central
Andes. The observed map relationships suggest no signif-
icant upper crustal deformation after 25 Ma, but do not
rule out the possibility of Neogene passive transport of
upper crustal material above deeper structures.
[85] 3. As a measure of Cenozoic exhumation, the
moderate-temperature (120–350°C) cooling history is de-
termined by a combination of fission track,
40
Ar/
39
Ar, and
MDD modeling results for 18 granite samples. The
40
Ar/
39
Ar results for muscovite, biotite, and K-feldspar
reveal a 45–40 Ma inception of rapid cooling, consistent
with new and previous ZFT data [Benjamin et al., 1987].
Modeling of K-feldspar and AFT results indicate that rapid
cooling continued throughout the late Eocene and Oligocene,
with a reduction in cooling rates at 26 Ma. These results
agree with the structural record, in which most shortening
preceded the Neogene. The erosional record of Eocene –
Oligocene exhumation is preserved in the eastern Altiplano
basin, where 8
40
Ar/
39
Ar tuff ages and sediment provenance
data reveal sedimentation between 33 and 25 Ma.
[86] 4. AFT thermochronology and thermal modeling for
16 granite samples constrain the low-temperature (60–
120°C) cooling history. Following an early to middle Mio-
cene period of isothermal or slow cooling conditions, a phase
of rapid cooling commenced at 11 Ma. This young cooling
episode postdated the dominantly pre-Neogene upper crustal
shortening of the region, suggesting the potential role of
exhumation mechanisms other than upper crustal deforma-
tion. Although there are several possibilities, the absence of
rapid late Miocene cooling in the low-precipitation, low-
erosion segment of the Eastern Cordillera in southern Bolivia
at 21°S[Ege et al., 2003; Ege, 2004] suggests that the more
erosive, high-precipitation climate north of 17.5°S may have
driven youthful exhumation in the Cordillera Real.
[87]Acknowledgments. This research was supported by grants from
the Geological Society of America and Sigma Xi (awarded to Gillis), and
National Science Foundation grant EAR-0510441 (awarded to Horton).
The
40
Ar/
39
Ar analyses were performed with assistance from Ana Vucic.
Fission track analyses were conducted by Raymond Donelick. We appre-
ciate helpful discussions with Alan Clark, Nadine McQuarrie, Daniel
Stockli, Sohrab Tawackoli, Carmala Garzione, Gary Axen, An Yin, Holly
Caprio, and Alexander Robinson. Field logistical support was provided by
Sohrab Tawackoli, Pedro Churata, and Jaime Tito of Sergeotecmin (La
Paz). Reviews by Jason Barnes, Barbara Carrapa, and Associate Editor
Todd Ehlers improved the manuscript.
TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
20 of 22
TC6007
References
Allmendinger, R. W., T. E. Jordan, S. M. Kay, and B. L.
Isacks (1997), The evolution of the Altiplano-Puna
plateau of the central Andes, Annu. Rev. Earth Pla-
net. Sci.,25, 139– 174.
Anders, M. H., K. M. Gregory-Wodzicki, and
M. Spiegelman (2002), A critical evaluation of late
Tertiary accelerated uplift rates for the Eastern
Cordillera, central Andes of Bolivia, J. Geol.,
110 , 89– 100.
Arriagada, C., P. Roperch, P. R. Cobbold, and
C. Mpodozis (2005), Numerical plan view restora-
tion of the Bolivian orocline, paper presented at In-
ternational Symposium on Andean Geodynamics,
ORSTOM, Paris.
Baby, P., P. Rochat, G. Mascle, and G. He´rail (1997),
Neogene shortening contribution to crustal thicken-
ing in the back arc of the central Andes, Geology,
25, 883– 886.
Baxter, E. F., D. J. DePaolo, and P. R. Renne (2002),
Spatially correlated anomalous
40
Ar/
39
Ar ‘‘age’’
variations in biotites about a lithologic contact near
Simplon Pass, Switzerland: A mechanistic explana-
tion for excess Ar, Geochim. Cosmchim. Acta,66,
1067 – 1083.
Beck, S. L., and G. Zandt (2002), The nature of orogenic
crust in the central Andes, J. Geophys. Res.,
107(B10), 2230, doi:10.1029/2000JB000124.
Benjamin, M. T. (1986), Fission-track ages on some
Bolivian plutonic rocks: Implications for the Ter-
tiary uplift and erosion history of the Altiplano-Cor-
dillera Real, M.S. thesis, 58 pp., Dartmouth Coll.,
Hanover, N. H.
Benjamin, M. T., N. M. Johnson, and C. W. Naeser
(1987), Recent rapid uplift in the Bolivian Andes:
Evidence from fission-track dating, Geology,15,
680 – 683.
Carleson, W. D., R. A. Donelick, and R. A. Ketcham
(1999), Variability of apatite fission-track annealing
kinetics: I. Experimental results, Am. Mineral.,84,
1213 – 1223.
Carlier, G., G. Grandin, G. Laubacher, R. Marocco, and
F. Megard (1982), Present knowledge of the mag-
matic evolution of the Eastern Cordillera, Peru,
Earth Sci. Rev.,18, 253 – 283.
Crough, S. T. (1983), Apatite fission-track dating of
erosion in the eastern Andes, Bolivia, Earth Planet.
Sci. Lett.,64, 396– 397.
Dallmeyer, R. D., and T. Rivers (1983), Recognition of
extraneous argon components through incremental-
release
40
Ar/
39
Ar analysis of biotite and hornblende
across the Grenvillian metamorphic gradient in
southwestern Labrador, Geochim. Cosmchim. Acta,
47, 413– 428.
DeCelles,P.G.,andB.K.Horton(2003),Earlyto
middle Tertiary foreland basin development and
the history of Andean crustal shortening in Bolivia,
Geol. Soc. Am. Bull.,115 , 58– 77.
Donelick, R. A., R. A. Ketcham, and W. D. Carleson
(1999), Variability of apatite fission-track annealing
kinetics: II. Crystallographic orientation effects,
Am. Mineral.,84, 1224– 1234.
Dorbath, C., M. Granet, G. Poupinet, and C. Martinez
(1993), A teleseismic study of the Altiplano and
Eastern Cordillera in northern Bolivia: New con-
straints on a lithospheric model, J. Geophys. Res.,
98, 9825– 9844.
Ege, H. (2004), Exhumations- und Hebungs-
geschichte der zentralen Anden in Su¨dbolivien
(21°S) durch Spaltspur-Thermochronologie en
Apatit, Ph.D. thesis, 173 pp., Freie Univ.,
Berlin.
Ege, H., E. Sobel, V. Jacobshagen, E. Scheuber, and
D. Mertmann (2003), Exhumation history of the
central Andes of southern Bolivia by apatite fission
track dating, Rev. Tec. Yacimientos Petrol. Fis.
Bolivianos,21, 165 – 172.
Evernden, J. F., S. J. Kriz, and C. Cherroni (1977),
Potassium-argon ages of some Bolivian rocks,
Econ. Geol.,72, 1042– 1061.
Farrar, E., A. H. Clark, D. J. Kontak, and D. A. Archibald
(1988), Zongo – San Gaba´n zone: Eocene foreland
boundary of the central Andean orogen, northwest
Bolivia and southeast Peru, Geology,16, 55 – 58.
Fitzgerald, P. G. (1994), Thermochronologic constraints
on post-Paleozoic tectonic evolution of the central
Transantarctic Mountains, Antarctica, Tectonics,13,
818 – 836.
Fitzgerald, P. G., and A. J. W. Gleadow (1988), Fission-
track geochronology, tectonics and structure of the
Transantarctic Mountains in northern Victoria Land,
Antarctica, Chem. Geol.,73, 169– 198.
Fitzgerald, P. G., R. B. Sorkhabi, T. F. Redfield, and
E. Stump (1995), Uplift and denudation of the
Alaska Range: A case study in the use of apatite
fission track thermochronology to determine abso-
lute uplift parameters, J. Geophys. Res.,100,
20,175 – 20,191.
Foster, D. A., T. M. Harrison, P. Copeland, and M. T.
Heizler (1990), Effects of excess argon within large
diffusion domains on K-feldspar age spectra, Geo-
chim. Cosmchim. Acta,54, 1699– 1708.
Garzione, C. N., P. Molnar, J. C. Libarkin, and B. J.
MacFadden (2006), Rapid late Miocene rise of the
Bolivian Altiplano: Evidence for removal of mantle
lithosphere, Earth Planet. Sci. Lett.,241, 543 – 556,
doi:10.1016/j.epsl.2005.11.026.
Gillis, R. J., B. K. Horton, and M. Grove (2004), Ex-
humation history and basin development along the
eastern margin of the central Andean plateau, Boli-
via, Geol. Soc. Am. Abstr. Programs,36, 433.
Gregory-Wodzicki, K. M. (2000), Uplift history of the
central and northern Andes: A review, Geol. Soc.
Am. Bull.,112 , 1091 – 1105.
Green, P. F. (1986), On the thermo-tectonic evolution of
northern England: Evidence from fission track ana-
lysis, Geol. Mag.,153, 493– 506.
Gubbels, T. L., B. L. Isacks, and E. Farrar (1993),
High-level surfaces, plateau uplift, and foreland de-
velopment, Bolivian central Andes, Geology,21,
695 – 698.
Harrison, T. M., M. T. Heitzler, O. M. Lovera,
W. Chen, and M. Grove (1994), A chlorine disin-

TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
21 of 22
TC6007
fectant for excess argon released from K-feldspar
during step heating, Earth Planet. Sci. Lett.,123,
95 – 104.
Heinrich, S. M. (1988), Geology and geochronology of
the Zongo river valley, Cordillera Oriental, NW
Bolivia, M.S. thesis, 185 pp., Queen’s Univ., King-
ston, Ont., Canada.
Henry, S. G., and H. N. Pollack (1988), Terrestrial
heat flow above the Andean subduction zone in
Bolivia and Peru, J. Geophys. Res.,93, 15,153 –
15,162.
Horton, B. K. (1999), Erosional control on the geome-
try and kinematics of thrust belt development in the
central Andes, Tectonics,18, 1292 – 1304.
Horton, B. K. (2005), Revised deformation history of
the central Andes: Inferences from Cenozoic fore-
deep and intermontane basins of the Eastern Cordil-
lera, Bolivia, Tectonics,24, TC3011, doi:10.1029/
2003TC001619.
Horton, B. K., B. A. Hampton, and G. L. Waanders
(2001), Paleogene synorogenic sedimentation in
the Altiplano plateau and implications for initial
mountain building in the central Andes, Geol.
Soc. Am. Bull.,113 , 1387 – 1400.
Horton, B. K., B. A. Hampton, B. N. LaReau, and
E. Baldello´n (2002), Tertiary provenance history of
the northern and central Altiplano (central Andes,
Bolivia): A detrital record of plateau-margin tec-
tonics, J. Sediment. Res.,72, 711– 726.
Ingersoll, R. V., T. F. Bullard, R. L. Ford, J. P. Grimm,
J. D. Pickle, and S. W. Sares (1984), The effect of
grain size on detrital modes: A test of the Gazzi-
Dickinson point-counting method, J. Sediment. Pet-
rol.,54, 103 – 116.
Isacks, B. L. (1988), Uplift of the central Andean pla-
teau and bending of the Bolivian orocline, J. Geo-
phys. Res.,93, 3211– 3231.
Kay,R. F., B.J. MacFadden, R.H. Madden, H. Sandeman,
and F. Anaya (1998), Revised age of the Salla beds,
Bolivia, and its bearing on the age of the Deseadan
South American land mammal ‘‘age’’, J. Vertebrate
Paleontol.,18,189–199.
Kennan, L., S. Lamb, and C. Rundle (1995), K-Ar dates
from the Altiplano and Cordillera Oriental of Boli-
via: Implications for Cenozoic stratigraphy and tec-
tonics, J. S. Am. Earth Sci.,8, 163– 186.
Kennan, L., S. H. Lamb, and L. Hoke (1997), High-
altitude palaeosurfaces in the Bolivian Andes: Evi-
dence for late Cenozoic surface uplift, in Palaeo-
surfaces: Recognition, Reconstruction and
Palaeoenvironmental Interpretation, edited by
M. Widdowson, Geol. Soc. Spec. Publ.,120, 307–323.
Ketcham, R. A., R. A. Donelick, and W. D. Carleson
(1999), Variability of apatite fission-track annealing
kinetics: III. Extrapolation to geological time scales,
Am. Mineral.,84, 1235– 1255.
Ketcham, R. A., R. A. Donelick, and M. B. Donelick
(2000), AFTSolve: A program for multi-kinetic
modeling of apatite fission-track data, Geol. Mater.
Res.,2, 1– 32.
Kley, J., and C. R. Monaldi (1998), Tectonic shortening
and crustal thickness in the central Andes: How
good is the correlation?, Geology,26, 723 – 726.
Kontak,D.J.,E.Farrar,A.H.Clark,andD.A.
Arachibald (1990), Eocene tectono-thermal rejuve-
nation of an upper Paleozoic-lower Mesozoic terrane
in the Cordillera de Carabaya, Puno, southeastern
Peru, revealed by K-Ar and
40
Ar/
39
Ar dating, J. S.
Am. Earth Sci.,3, 231 – 246.
Lamb, S., and L. Hoke (1997), Origin of the high pla-
teau in the central Andes, Bolivia, South America,
Tectonics,16, 623 – 649.
Lamb, S., L. Hoke, L. Kennan, and J. F. Dewey (1997),
Cenozoic evolution of the Central Andes in Bolivia
and northern Chile, in Orogeny Through Time, edi-
ted by J. P. Burg, and M. Ford, Geol. Soc. Spec.
Publ.,121, 237 – 264.
Lancelot, J. R., G. Laubacher, R. Marocco, and
U. Renaud (1978), U/Pb radiochronology of two
granite plutons from the Eastern Cordillera (Peru):
Extent of Permian magmatic activity and conse-
quences, Geol. Rundsch.,67, 236 – 243.
Laslett, G. M., P. F. Green, I. R. Duddy, and A. J. W.
Gleadow (1987), Thermal annealing of fission
tracks in apatite 2. A quantitative analysis, Chem.
Geol.,65, 1 – 13.
Lee, J. K. W. (1995), Multipath diffusion in geochro-
nology, Contrib. Mineral. Petrol.,120, 60 – 82.
Lehmann, B. (1978), A Precambrian core sample from
the Altiplano/Bolivia, Geol. Rundsch.,67, 270 –
278.
Lovera,O.M.,F.M.Richter,andT.M.Harrison
(1989), The
40
Ar/
39
Ar thermochronometry for
slowly cooled samples having a distribution of dif-
fusion domain sizes, J. Geophys. Res.,94, 17,917 –
17,935.
Lovera, O. M., M. Grove, T. M. Harrison, and K. I.
Mahon (1997), Systematic analysis of K-feldspar
40
Ar/
39
Ar step heating r esults: I. Significanc e of
activation energy determinations, Geochim. Cosm-
chim. Acta,61, 3171 – 3192.
Lovera, O. M., M. Grove, T. M. Harrison, and K. I.
Mahon (2002), Systematic analysis of K-feldspar
40
Ar/
39
Ar step heating results: II. Relevance of la-
boratory argon diffusion properties to nature, Geo-
chim. Cosmchim. Acta,66, 1237 – 1255.
Lutz, T. M., and G. Omar (1991), An inverse method of
modeling thermal histories from apatite fission-
track, Earth Planet. Sci. Lett.,104, 181 – 195.
MacFadden, B. J., K. E. Campbell, R. L. Cifelli,
O. Siles, N. M. Johnson, C. W. Naeser, and P. K.
Zeitler (1985), Magnetic polarity stratigraphy and
mammalian fauna of the Deseadan (late Oligo-
cene –early Miocene) Salla Beds of northern Boli-
via, J. Geol.,93, 223 – 250.
Marsh, S. P., D. H. Richter, S. Ludington, E. Soria-
Escalante, and A. Escobar-Diaz (1992), Geologic
map of the Altiplano and Cordillera Occidental,
Bolivia, in Geology and mineral resources of the
Altiplano and Cordillera Occidental, Bolivia:
U.S.G.S. and Servicio Geolo´gico de Bolivia, U.S.
Geol. Surv. Bull.,1975, plate 1.
Martinez, C. (1980), Structure et e´ volution de la
chaıˆne hercynienne et da la cha
˜circ;ne andine dans
le nord de la Cordille`re des Andes de Bolivie, Tr av.
Doc. ORSTOM,119 , 352 pp.
Masek, J. G., B. L. Isacks, T. L. Gubbels, and E. J.
Fielding (1994), Erosion and tectonics at the mar-
gins of continental plateaus, J. Geophys. Res.,99,
13,941 – 13,956.
McBride, S. L., R. C. R. Robertson, A. H. Clark, and
E. Farrar (1983), Magmatic and metallogenetic
episodes in the northern tin belt, Cordillera Real,
Bolivia, Geol. Rundsch.,72, 685 – 713.
McBride, S. L., A. H. Clark, E. Fa rrar, and D. A.
Archibald (1987), Delimitation of a cryptic Eocene
tectono-thermal domain in the Eastern Cordillera
of the Bolivian Andes through K-Ar dating and
40
Ar-
39
Ar step-heating, J. Geol. Soc. London,
144, 243 – 255.
McDougall, I., and T. M. Harrison (1999), Geochronol-
ogy and Thermochronology by the
40
Ar/
39
Ar Method,
2nd ed., 269 pp., Oxford Univ. Press, New York.
McQuarrie, N. (2002), The kinematic history of the
central Andean fold-thrust belt, Bolivia: Implica-
tions for building a high plateau, Geol. Soc. Am.
Bull.,114 , 950 – 963.
McQuarrie, N., and P. G. DeCelles (2001), Geometry
and structural evolution of the central Andean back-
thrust belt, Bolivia, Tectonics,20, 669 – 692.
McQuarrie, N., B. K. Horton, G. Zandt, S. Beck, and P. G.
DeCelles (2005a), Lithospheric evolution of the An-
dean fold-thrust belt, Bolivia, and the origin of the
central Andean plateau, Tectonophysics,399,15–
37.
McQuarrie, N., J. B. Barnes, and T. A. Ehlers (2005b),
The relationship between deformation and exhuma-
tion in the central Andean plateau, Bolivia, Geol.
Soc. Am. Abstr. Programs,37, 271.
McRae, L. E. (1990), Paleomagnetic isochrons, un-
steadiness, and uniformity of sedimentation in
Miocene intermontane basin sediments at Salla,
eastern Andean Cordillera, Bolivia, J. Geol.,98,
479 – 500.
Miller, J. F. (1988), Granite petrogenesis in the Cordil-
lera Real, Bolivia and crustal evolution in the cen-
tral Andes, Ph.D. thesis, 316 pp., Open Univ.,
London.
Miller, J. F., and N. J. Harris (1989), Evolution of
continental crust in the central Andes: Constraints
from Nd isotope systematics, Geology,17, 615 –
617.
Montgomery, D. R., G. Balco, and S. D. Willett (2001),
Climate, tectonics, and the morphology of the An-
des, Geology,29, 579 – 582.
Moore, M. A., and P. C. England (2001), On the
inference of denudation rates from cooling ages
of minerals, Earth Planet. Sci. Lett.,185, 265 –
284.
Newell, N. D. (1949), Geology of the Lake Titicaca
region, Peru and Bolivia, Mem. Geol. Soc. Am.,
36, 111 pp.
Pareja, J., C. Vargas, R. Sua´rez, R. Ballo´ n, R. Carrasco,
and C. Villaroel (1978), Mapa geolo´gico de Bolivia
y memoria explicative, scale 1:1,000,000, 27 pp.,
Yacimientos Petrol. Fis. Bolivianos and Serv. Geol.
de Bolivia, La Paz.
Parsons,I.,W.L.Brown,andJ.V.Smith(1999),
40
Ar/
39
Ar thermochronology using alkali feldspars:
Real thermal history or mathematical mirage of mi-
crotexture?, Contrib. Mineral. Petrol.,136,92–
110.
Roddick, J. C. (1978), The application of isochron dia-
grams in
40
Ar-
39
Ar dating: A discussion, Earth Pla-
net. Sci. Lett.,41, 233 – 244.
Roddick, J. C., R. A. Cliff, and D. C. Rex (1980), The
evolution of excess argon in Alpine biotites - An
40
Ar-
39
Ar analysis, Earth Planet. Sci. Lett.,48,
185 – 208.
Roeder, D. (1988), Andean-age structure of Eastern Cor-
dillera (province of La Paz, Bolivia), Tec tonic s,7,
23 – 39.
Safran, E. B. (1998), Channel network incision and
patterns of mountain geomorphology, Ph.D. thesis,
326 pp., Univ. of Calif., Santa Barbara.
Sandeman, H. A., A. H. Clark, and E. Farrar (1995),
An integrated tectono-magmatic model for the
evolution of the southern Peruvian Andes (13 –
20°S) since 55 Ma, Int. Geol. Rev.,37, 1039 –
1073.
Santos, J. O. S., 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 on integration of field
mapping and U-Pb and Sm-Nd geochronology,
Gondwana Res.,3, 453 – 488.
Sempere, T., G. He´rail, J. Oller, and M. G. Bonhomme
(1990), Late Oligocene – early Miocene major tec-
tonic crisis and related basins in Bolivia, Geology,
18, 946 – 949.
Servicio Geolo´gico de Bolivia (Geobol) (1993), Carta
geologica de Bolivia, Ichoca (Hoja 6142), scale
1:100000, La Paz.
Servicio Geolo´gico de Bolivia (Geobol) (1995), Carta
geologica de Bolivia, Milluni (Hoja 5945), scale
1:100000, La Paz.
Servicio Geolo´gico de Bolivia (Geobol) (1997), Carta
geologica de Bolivia, Inquisivi (Hoja 6143), scale
1:100000, La Paz.
Servicio Nacional de Geologı´a y Te´ cnico de Minas and
Yacimientos Petrolı´feros Fiscales Bolivianos
(2003), Mapa geologico de Bolivia, scale
1:100,000, La Paz.
Sheffels, B. M. (1990), Lower bound on the amount of
crustal shortenin g in the central Bolivian Andes,
Geology,18, 812 – 815.
Stu¨we, K., L. White, and R. Brown (1994), The influ-
ence of eroding topography on steady state iso-
therms: Application to fission track analysis,
Earth Planet. Sci. Lett.,124, 63 – 74.
Sua´rez, R., and E. Diaz (1996), Lexico estratigrafico de
Bolivia, Rev. Te´c. Yacimientos Petrol. Fis. Bolivia-
nos,17(1 – 2), 7 – 227.
Willett, S. D. (1997), Inverse modeling of annealing of
fission tracks in apatite. 1: A controlled random
search method, Am. J. Sci.,297, 939 – 969.

TC6007 GILLIS ET AL.: EXHUMATION OF THE CENTRAL ANDEAN PLATEAU
22 of 22
TC6007
Wo¨ rner, G., J. Lezaun, A. Beck, V. Heber, F. Lucassen,
E. Zinngrebe, R. Ro¨ssling, and H. G. Wilke (2000),
Precambrian and Early Paleozoic evolution of the
Andean basement at Belen (northern Chile) and
Cerro Uyarani (western Bolivia Altiplano), J. S.
Am. Earth Sci.,13, 717 – 737.
Zeitler, P. K., and J. D. Fitz Gerald (1986), Saddle-
shaped
40
Ar/
39
Ar spectra from young, microstruc-
turally complex potassium feldspars, Geochim.
Cosmchim. Acta,50, 1185 – 1199.

R. J. Gillis, Division of Geological and Geophysical
Surveys, State of Alaska Department of Natural
Resources, 3354 College Road, Fairbanks, AK 99709-
3707, USA. (robert_gillis@dnr.state.ak.us)
M. Grove, Department of Earth and Space Sciences,
University of California, 595 Charles Young Drive East,
3806 Geology Building, Los Angeles, CA 90095-1567,
USA.
B. K. Horton, Department of Geological Sciences,
Jackson School of Geosciences, University of Texas, 1
University Station C1100, Austin, TX 78712-0254,
USA.