Variation MagmaComposit CentralAndes MMamani

Variations in magma composition in time and
space along the Central Andes (13°S-28°S)
Dissertation
zur Erlangung des Doktorgrades

der Mathematisch-Naturwissenschaftlichen Fakultäten
der Georg-August-Universität zu Göttingen
vorgelegt von
Mirian-Irene Mamani-Huisa
aus Cuyocuyo (Peru)
Göttingen 2006
D7
Referent: Prof. Dr. G. Wörner

Koreferent: Prof. Dr. B.T. Hansen
Tag der mündlichen Prüfung: 24, Oktober 2006
1
Contents
Abstract....................................................................................................................................... 3
Zusammmenfassung ................................................................................................................... 5
Resumen ..................................................................................................................................... 7
Acknowledgements .................................................................................................................... 9
Preamble ................................................................................................................................... 10
1 Introduction ........................................................................................................................... 11
1.1 Variations in magma composition in time and space along the Central Andes (13°S-
28°S): Facts and open questions........................................................................................... 11
1.2 Working Hypothesis ....................................................................................................... 11
1.3 Previous work ................................................................................................................. 12
1.3.1 The Andes................................................................................................................12
1.3.2 Pre-Andean history .................................................................................................. 15
1.3.3 The Andean Cycle ................................................................................................... 17
1.3.4 Uplift of the Central Andes and subducting Nazca plate ........................................18
1.3.5 Andean volcanism ................................................................................................... 20
1.3.6 Magma genesis ........................................................................................................ 21
2 Geochemical variations in south Peruvian volcanic rocks (13°S-18°S): The role of crustal
composition and thickening through time and space ............................................................... 25
2.1 Introduction .................................................................................................................... 25
2.2 Tectonic Setting.............................................................................................................. 26
2.3 Description of Cenozoic units and sampled volcanic centers ........................................27
2.3.1 Middle Eocene - Lower Oligocene, ~ 45 to ~30 Ma (Lower Moquegua Formation,
Anta Group)...................................................................................................................... 27
2.3.2 Upper Oligocene - Middle Miocene, ~30 to ~15 Ma (Upper Moquegua Formation,
Tacaza Group) .................................................................................................................. 28
2.3.3 Upper Miocene – Lower Pliocene, ~10 to ~3 Ma (Lower Barroso Formation)...... 29
2.3.4 Pliocene, ~3 to ~2 Ma (Sencca Formation) ............................................................. 29
2.3.5 Upper Pliocene to Pleistocene, 3 to 0.8 Ma (Upper Barroso Formation)................29
2.3.6 Holocene Volcanoes < 0.01 Ma .............................................................................. 29
2.4 Results ............................................................................................................................ 30
2.4.1 Geochemistry........................................................................................................... 30
2.4.1.1 Temporal variations of element ratios .............................................................. 39
2.4.1.2 Isotopes............................................................................................................. 40
2.5 Discussion....................................................................................................................... 45
2.5.1 Chemical isotopic composition of magmatic rocks in southern Peru and northern
Chile ................................................................................................................................. 46
2.5.2 Lithopheric Cross Sections and the Tectonic and Magmatic History in southern
Peru................................................................................................................................... 52
2.6 Conclusions .................................................................................................................... 56
3 Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the Central Andes
(13°S to 28°S)........................................................................................................................... 57
3.1 Introduction .................................................................................................................... 57
3.2 Database ......................................................................................................................... 58
2
3.3 Geochemistry of Meso-Cenozoic magmatic rocks in the CA ........................................ 60

3.3.1 Major and trace elements......................................................................................... 60
3.3.2 Isotopes.................................................................................................................... 66
3.4 Summary and Discussion: Implications for a tectonic model ........................................ 71
3.4.1 Source or Processes? ............................................................................................... 72
3.4.2 Geochemical constraint in a Tectonic model ..........................................................80
4 Geochemical variations in time and space of lead isotopic domains in the Central Andes
(13°S -28°S): Implications of the crustal structure and metal sources..................................... 82
4.1 Introduction .................................................................................................................... 82
4.2 Information systems and Pb isotope data .......................................................................83
4.2.1 Lead isotope data ..................................................................................................... 83
4.3 Results ............................................................................................................................ 84
4.3.1 Mapping of Pb isotopes and geochemical variations in Pb-domains ...................... 84
4.4 Discussion....................................................................................................................... 87
4.4.1 Pb isotope mapping crustal domains ....................................................................... 87
4.4.1.2 Isotopic framework of basement in the Central Andes .................................... 88
4.4.2 Crustal Contamination in the Pb domains ...............................................................92
4.4.3 Isotopic domains constrained by 3D density model................................................ 94
4.4.4 The Nature of the Transition Zones......................................................................... 95
4.4.5 Segmentation of the Central Andes plateau related to the Pb-isotope domains...... 96
4.4.6 Rheological and structural identity of the Pb domains during Andean Orogeny.... 97
4.4.7 Pb-isotopes of gold, silver copper and tin deposits, their relationship with the Pb-
domains............................................................................................................................. 98
4.5 Conclusions .................................................................................................................. 103
5 Appendix ............................................................................................................................. 104

5.1 Sample locations...........................................................................................................104
5.2 Analytical Methods ...................................................................................................... 105
5.2.1 Sample Preparation................................................................................................ 105
5.2.2 XRF (X-Ray Fluorescence Spectroscopy) analysis .............................................. 106
5.2.3 ICP-MS (Inductively Coupled Plasma Mass Spectrometry) analysis ................... 106
5.2.4 TIMS (Thermal Ionization Mass Spectrometer) analysis .....................................107
6 References ........................................................................................................................... 108

3
Abstract
This dissertation investigates the diversity of magma chemistry along and across the arc with
time in the Central Andes, with emphasis in the Neogene magmatism. The main body of this
thesis is formed by three independent chapters. The specific regions, results and conclusions
of these individual works can be summarized as follow:
Chapter 2: Geochemical variations in south Peruvian volcanic rocks (13°S-18°S): The role of
crustal composition and thickening through time and space
Chemical and isotopic data from 36 volcanic centers (from Eocene to Holocene) of the
northern Central Volcanic Zone (CVZ) in southern Peru. Volcanic rocks of these centers
record the beginning as well as the peak of shortening and crustal thickening in the Upper
Oligocene-Lower Miocene. Samples show systematic chemical and isotopic differences with
age. The compositions of andesites erupted before and after crustal thickening are similar in
terms of major elements. However, Post-Miocene andesites show enrichment in trace
elements (Ba, Sr), LREE (La, Sm) as well as high 87Sr/86Sr and low εNd. This indicates greater
crustal contamination compared to the older equivalents. Pb-isotopic ratios behave differently,
are mostly (25-0 Ma) independent of age, but change abruptly at 16°S, and in any given sector
206
Pb/204Pb ratios are similar. Comparison of contamination indicators with age show that
contamination was low in the Tacaza arc (30 Ma to 15 Ma), increased sharply in the Lower
Barroso arc (10 to 3 Ma), and remained at a high level up to the Present frontal-arc since then.
These younger volcanoes (< 3 Ma) show large ranges of Sm/Yb ratios (e.g. Sara Sara,
Huaynaputina).
I conclude that Y and HREE depletion from Miocene to Pleistocene volcanic rocks is caused
by residual garnet of crustal assimilation after crustal thickening. Any involvement of slab
melts in northern CVZ rocks to explain the “adakitic” signature is also excluded. The existing
variations in lead isotopes in samples of similar age at 16°S support the notion that
geochemical compositions in the magmas are controlled by the composition of the underlying
basement.
Chapter 3: Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the

Central Andes (13°S to 28°S)
Meso-Cenozoic magmatic arc systems of the Central Andes (CA) result from the subduction
of the Nazca Plate beneath the South America plate since Jurassic time, the arc progressively
shifted ~180 km from a western position in the Jurassic to the present Western Cordillera in
the Oligocene time. These magmas formed prior, during and after crustal thickening. Present-
day continental crust reaches a thickness of >70 km. Enhanced uplift between c. 30 and 15
Ma is documented by large clastic wedges in Western Andean Escarpment. The upper crust is
formed by Precambrian to Paleozoic metamorphic rocks and sedimentary marine Mesozoic
rocks, covered by mostly continental sediments of the Cretaceous and Tertiary.
Evolutionary pattern in geochemistry of rocks younger than mid-Tertiary show no evidence of
crustal interaction involving garnet, in accordance with a thin crust at these times. While all
older rocks are low in Sm/Yb, younger rocks (< 3 Ma) may be both, high and low in these
ratios. Thus, high Sm/Yb in a particular rock cannot be simply taken as a proxy for thick
crust.
Pliocene-Holocene volcanics north of 22°S show stronger indication of interaction with lower
garnetiferous crust. To the South Sm/Yb is lower although crustal thickness is similar. This
difference may then reflect the additional effect of variable crustal composition, north and
south of 22°S. Significant isotopic and trace element differences are observed for rocks of the
same age but at slightly different locations. Regional differences exist in other trace element
ratios as well. This supports the notion that the crust not only controls the isotopes
4
composition of magmas but also their trace element patterns. Thus, caution needs to be
applied when plotting geochemical data vs. age for rocks from a wide regional distribution as
local basement control may be more important than age in controlling their composition.
Chapter 4: Geochemical variations in time and space of Lead isotopic domains in the Central
Andes (13°S -28°S): Implications of the crustal structure and metal sources
Lead isotope compositions of 802 Pb isotope analysis (356 previously published, 446 new),
Nd-Sr isotope values and trace elements (150 published data, 180 new) on Proterozoic to
Holocene igneous, metamorphic, and sedimentary rocks as well as Jurassic to Mio-Pliocene
arc-related ore deposits define - at high spatial resolution - distinct isotopic domains of the
crust in the Central Andes. These domains correlate with crustal structure index (Ө) mapping
based on geometries of the 3D density model. Pb isotopic boundaries thus correspond to
variations in the intracrustal density structure that reflects distinct mafic and felsic crustal
compositions. Our combined isotopic and geophysical mapping suggests that the variations in
crustal composition must be old and that crustal evolution as reflected in its structure, age and
composition had an important control on major element heterogeneity (mafic versus felsic),
lead and Neodymium isotopes heterogeneity and Central Andean plateau segmentation.
5
Zusammmenfassung
Diese Dissertation untersucht die Diversität der Magmenchemie entlang sowie quer des
magmatischen Bogens in Bezug auf die temporären Veränderungen in den zentralen Anden,
mit Schwerpunkt auf dem Neogenen Magmatismus. Der Hauptteil dieser Arbeit ist in drei
unabhängige Kapitel gegliedert. Die Spezifischen Regionen, Ergebnisse und Rückschlüsse
dieser drei individuellen Kapitel können wie folg zusammengefasst werden:
Grundlage sind chemische und isotopische Analysen von 36 vulkanischen Zentren (vom
Eozän bis Holozän) der nördlichen Central Volcanic Zone (CVZ) in Süd Peru. Die Vulkanite
dieser Zentren spiegeln den Beginn und auch den Höhepunkt der Verkürzung und der damit
assoziierten Krustenverdickung vom oberen Oligozän bis unteres Miozän wider. Die Proben
zeigen systematische Veränderungen in der Chemie und Isotopie in Abhängigkeit von der
Zeit. Die Zusammensetzungen von Andesiten die vor und nach der Krustenverdickung
eruptiert wurden sind bezüglich der Hauptelemente ähnlich. Dahingegen zeigen post Miozäne
Andesite Anreicherung an bestimmten Spurenelementen (Ba, Sr), LSEE (La, Sm) sowie auch
hohe 87Sr/86Sr und niedrige εNd Werte. Dies sind Indikatoren für größere krustale
Kontamination im Vergleich zu den älteren Äquivalenten. Die Pb-Isotopenverhältnisse
verhalten sich anders, sie sind zumeist (25-0 Ma) unabhängig vom Alter, verändern sich aber
plötzlich bei 16°S und in bestimmten definierten Regionen sind die 206Pb/204Pb Verhältnisse
ähnlich. Vergleiche von Kontaminationsindikatoren in Bezug auf das Alter zeigen, dass die
Kontamination gering war während des Tacaza Bogens (30 to 15 Ma), stark anstieg während
des unteren Barroso Bogens (10 to 3 Ma) und auf einem hohen Level seit dieser Zeit blieb bis
zum heutigen Frontalen Bogen. Diese jungen Vulkanite (< 3 Ma) zeigen ein großes Spektrum
in Sm/Yb Verhältnissen (e.g. Sara Sara, Huaynaputina).
Ich schließe daraus, dass die Verarmung von Y und den SSEE in den Vulkaniten vom Miozän
zum Pleistozän durch Granat im Residuum bei der Assimilation nach der Krustenverdickung
entstanden ist. Des Weiteren lässt sich die „Adakit-Signatur“ von Gesteinen der nördlichen
CVZ nicht durch Schmelzen aus der subduzierten Platte erklären. Die gefundenen Variationen
in der Blei Isotopie in Proben gleichen Alters in der Region um 16°S erhärtet die Auffassung,
dass die geochemische Zusammensetzung der Magmen durch die Zusammensetzung des
darunter liegenden Grundgebirge kontrolliert wird.

Das Meso-Känozäne magmatische Bogen System der Zentralen Anden (CA) wird
hervorgerufen durch die Subduktion der Nazca Platte unter die Südamerikanische Platte seit
dem Jura. Der Bogen bewegte sich kontinuierlich ~ 180 km von einer westliche Position im
Jura hin zur heutigen westlichen Gebirgskette im Oligozän. Diese Magmen wurden früher
gebildet, während und nach der Krustenverdickung. Die heutige kontinentale Kruste erreicht
eine Dicke von >70 km. Erhöhte Hebung zwischen ca. 30 und 15 Ma ist dokumentiert durch
große klastische Keile in den Schichtstufen der westlichen Anden. Die obere Kruste besteht
aus Präkambrischen und Paläozoischen Metamorphiten sowie Mesozänen marinen
Sedimenten, die meist mit kontinentalen Sedimenten aus der Kreide und dem Tertiär bedeckt
sind.
Die Entwicklugsmuster in der Geochemie der Gesteine jünger als mittel Tertiär zeigen keine
Beweise für Krusteninteraktion mit Granateinwirkung, was in Übereinstimmung mit der
dünnen Kruste zu dieser Zeit ist. Während alle älteren Gesteine ein niedriges Sm/Yb
6
Verhältnis haben, sind jüngere Gesteine (< 3 Ma) variabel und zeigen beides, hohe und
niedrige Werte. Folglich können hohe Sm/Yb Verhältnisse in einzelnen Gesteinen nicht
einfach als Indikatoren für eine mächtige Kruste herangezogen werden.
Pliozäne-Holozäne Vulkanite nördlich 22°S zeigen stärkere Anzeichen für Einwirkungen von
tiefer granitischer Kruste. Im Süden ist das Sm/Yb Verhältnis geringer, obwohl die
Krustendicke identische ist. Dieser Unterschied spiegelt dann den zusätzlichen Effekt der
variablen Krustenzusammensetzung im Norden und südliche 22°S wider. Signifikante
Unterschiede in Isotopen und Spurenelement werden in Gesteinen gleichen Alters beobachtet,
unterscheiden sich aber etwas in den unterschiedlichen Regionen. Regionale Unterschiede
existieren auch in anderen Spurenelementverhältnissen. Dies unterstützt die Annahme, dass
die Zusammensetzung der Kruste nicht nur die Isotopenzusammensetzung sondern auch die
Spurenelementmuster der Magmen kontrolliert. Folglich muss große Vorsicht gewährt
werden wenn die geochemischen Daten gegen das Alter für bestimmte Gesteine mit einer
großen regionalen Verteilung aufgetragen werden, denn das lokale Grundgebirge könnte die
Zusammensetzung stärker beeinflussen als das Alter.
Die Blei Isotopen Zusammensetzung von 802 Pb-Isotopen Analysen (356 bereits
veröffentliche, 446 neue) sowie die Nd-Sr Werte und Spurenelementdaten (150 bereits
veröffentliche, 180 neue) an Proterozoischen bis Holozänen magmatischen, metamorphen und
sedimetätren Gesteinen wie auch an Jurassischen bis Mio-Pliozänen Bogen assoziierten
Erzlager definieren - bei hoher räumlicher Auflösung – eindeutige Isotopendomänen der
Kruste in den Zentral Anden. Diese Domänen korrelieren mit der Krusten-Struktur-Index (Ө)
Kartierung welche auf den Geometrien einer 3D Dichte Modellierung basiert. Die Grenzen
der Pb Isotope entsprechen den Variationen in den intrakrustalen Dichte Strukturen, welche
mafische und felsische Krustenzusammensetzungen widerspiegeln. Unsere kombinierte
isotopische und geochemische Kartierung weist darauf hin, dass die Variationen in der
krustalen Zusammensetzung alt sind. Weiterhin hatten die Krustenentwicklung, die in diesen
Strukturen festgehalten ist, sowie Alter und Zusammensetzung einen großen Einfluss auf die
Heterogenität der Hauptelemente (mafisch gegen felsisch), die Heterogenitäten der Blei und
Neodym Isotope, und die Segmentierung der Zentral Anden Hochebene.
7
Resumen
Esta disertación investiga la diversidad de la química en espacio y tiempo de los magmas en

los arcos de los Andes Centrales, con énfasis en el magmatismo del Neógeno. El principal
contenido de esta Tesis esta conformado por tres independientes capítulos. Las regiones
específicas, resultados y conclusiones de estos trabajos individuales pueden ser resumidos
como sigue:
Datos químicos y isotópicos de 36 centros volcánicos (desde el Eoceno hasta el Holoceno) del
norte de la Zona Volcánica Central (ZVC) en el sur de Perú. Las rocas volcánicas de estos
centros registran el comienzo y el máximo acortamiento, y espesor cortical en el Oligoceno
superior-Mioceno inferior. Las muestras indican diferencias en la sistemática de la química y
isótopos con la edad. La composición de las andesitas eruptadas antes y después del
espesamiento cortical son similares en términos de elementos mayores. Sin embargo, las
andesitas post-Mioceno muestran enriquecimiento en elementos traza (Ba, Sr), LREE (La,
Sm) así como también alto 87Sr/86Sr y bajo εNd. Esto indica gran contaminación cortical
comparado con los equivalentes antiguos.
Las relaciones de isótopos de Pb son diferentes, mayormente independientes de la edad (25-0
Ma), y cambian abruptamente a 16°S, y en determinados sectores las relaciones 206Pb/204Pb
son similares. Las comparaciones de indicadores de contaminación con la edad muestran que
la menor contaminación fue en el arco del Tacaza (20 Ma to 10 Ma), incremento fuertemente
en el arco del Barroso inferior (10 to 3 Ma), y permanece desde entonces a mayor nivel en el
presente arco frontal. Estas lavas recientes (< 3 Ma) muestran amplios rangos en relaciones de
Sm/Yb (e.g. Sara Sara, Huaynaputina).
Las depleciones de Y y HREE desde el Mioceno hasta el Pleistoceno fueron causados por
granate residual de la asimilación cortical después del espesamiento de la corteza. Se excluye
algún envolvimiento de fusión de la corteza subductada en las rocas del norte de la ZVC para
explicar las signaturas de “adakitas”. Las variaciones existentes de isótopos de plomo en las
muestras de similar edad a 16°S soporta la idea que la composición geoquímica en los
magmas fueron controlados por la composición del basamento infrayaciente.

El sistema de arcos magmáticos Meso-Cenozoicos en los Andes Centrales resulta de la
subducción de la placa de Nazca por debajo de la placa de Sud-América desde el Jurásico. El
arco progresivamente se desplazo ~180 km desde una posición al oeste en el Jurásico hasta
su posición actual en la Cordillera Occidental en el Oligoceno. Estos magmas se formaron
antes, durante y después del espesamiento cortical. Actualmente el espesor de la corteza
continental alcanza un espesor >70 km. El aumento del levantamiento entre c. 30 y 15 Ma
esta documentado por la larga sedimentación en la pendiente oeste de la Cordillera
Occidental.
La corteza superior esta formada por rocas metamórficas del Proterozoico y Paleozoico y
sedimentos marinos del Mesozoico, cubiertos en su mayoría por sedimentos continentales del
Cretácico y Terciario. El patrón evolucionarío de la geoquímica de las rocas anteriores al
Terciario-medio no muestra evidencia de interacción cortical envolviendo granate, en acuerdo
con una corteza delgada en esos tiempos. Mientras que todas las rocas antiguas son menores
en las relaciones de Sm/Yb, las rocas recientes (< 3 Ma) tienen ambos, mayores y menores.
8
Estas relaciones mayores de Sm/Yb en una roca particular no pueden ser tomada simplemente
como una medida de corteza espesa.
Volcanicos del Plioceno-Holoceno al norte de 22°S muestran una indicación fuerte de
inteacción con corteza inferior granatifera. Hacia el sur Sm/Yb is menor a pesar de que el
espesor de la corteza es similar. Esta diferencia al norte y sur de 22°S puede luego reflejar el
efecto adicional en la variación de la composición de la corteza. Significantes diferencias en
isótopos y elementos trazas son observados en rocas de igual edad pero en pequeñas
locaciones diferentes. Diferencias regionales existen también en otras relaciones de elementos
trazas. Esto soporta la idea que la corteza no solo controla la composición isotópica de los
magmas, tambien controla sus patrones de elementos traza. Por eso, se necesita cuidado
cuando se usan ploteos de datos geoquímicas vs. la edad de las rocas de una distribución
regional extensa ya que el control del basamento local puede ser mas importante que la edad
en controlar su composición.
Composiciones de isótopos de Pb analizadas en 802 muestras (356 previamente publicadas,
446 nuevas) de rocas ígneas y sedimentarias del Proterozoico al Holoceno, así como también
minerales de depósitos relacionados a arcos desde el Jurasico al Mio-Plioceno - definen a alta
resolución - distintos dominios isotópicos de la corteza en los Andes Centrales. Estos
dominios se correlacionan con discontinuidades de densidad dentro de la corteza mapeados
sobre la base de la geometría en la estructura intracortical. Los límites de los isótopos de Pb
corresponden a la variación en la estructura de la densidad cortical que refleja distintas
composiciones corticales como máficas y félsicas. Nuestro mapeo combinado de isótopos y
geofísica sugiere que las variaciones en la composición cortical deben ser antiguas, y la
evolución intracortical como es reflejado en su estructura, edad y composición tienen un
importante control en la heterogeneidad de elementos mayores (máficos versus félsicos),
heterogeneidad en los isótopos de plomo y neodimio, y segmentación de las altiplanicies en
los Andes Centrales.
9
Acknowledgements
I wish to give my best thanks to my advisor Prof. Gerhard Wörner. He kindly opened the
possibility to come to Germany, supported the application to the DAAD scholarship, gave
from the beginning all that I needed to develop my work and scientific basis and skills he
imparted to me to understand the geochemistry of the igneous rocks, its compositions,
interpretation and relation with other geology branch. We spend many hours together
discussing the multifarious aspects of my research, and he always made time for my
questions.
I am grateful to the Deutscher Akademischer Austauschdienst (DAAD) for the scholarship I
received to develop my doctoral thesis in Germany. The Geowissenschaftliches Zentrum
Göttingen (GZG) at the Georg August Universität Göttingen welcomed me.
I am very grateful to Gerald Hartmann who introduced me to ion exchange column separation
and carried out the TIMS analysis. I also thank Klaus Simon who spent many hours to make
the ICP MS measurements.
I am grateful to the technical group of the Geochemistry Departament, from which I received
assistance during the laboratory work. From this group I wish to thank Ingrid Reuber, Erwin
Schiffczyk and Angelika Reitz and essential tips in the handling of computer I thank Stefan
Möller.
Occasional rich discussions, inspiring ideas on field trips or at congresses, with T. Sempere,
D. Cassard, A. Tassara, F. Lucassen, M. Fornari and J.C. Thouret had a positive influence on
the course of the present work and are gratefully acknowledged.
I would like to thank the people of the GIS Andes Project of the French Geologycal Survey
(BRGM), for his helpful advice in the GIS programs. I also thank Yves Husson for help me
with database management and processing.
I thank the members of the Peru Geological Survey (INGEMMET) for their cooperation.
I sincerely wish to acknowledge the generosity, friendly help, collaboration and constructive
discussions of Philipp Ruprecht, Birgit Scheibner, Aneta Kiebala, Silvia Rosas, Wenke
Wegner, Magdalena Banaszak, Silke Triebold, Andreas Kronz, Arnd Heumann, Janina Klaus
and Sonja Pabst.
Finally, I wish to thank my mother Basilia Huisa, my father Carlos Mamani, my sisters Nora
and Marisol, my brothers Carlos and Joel and the whole family in Peru. They have been the
spiritual force that enabled me to meet this challenge, thus expanding my knowledge and skill
to better research. Although we have not been physically close during my residence in
Germany, I have always felt that they are by my side.
This dissertation and all the efforts that came along with it are dedicated to the “Pachamama”!
10
Preamble
This dissertation studies the geochemistry and isotope composition of Meso-Cenozoic

magmas formed before, during and after crustal thickening in the Central Andes that serves as
a tool to analyse the causes of their variations in space and time. It consists of six chapters.
Chapter 1 is an introduction to the dissertation discussing the open questions related

with the geochemical variation of volcanic rocks in the Central Andes that are intented to be
addressed throughout this work.
This Chapter also summarises the main hypothesis driving the scientific work and previous
work.
Chapter 2 is titled “Geochemical variations in south Peruvian volcanic rocks (13°S-
18°S): The role of crustal composition and thickening through time and space”. I present the
results of new geochemical analysis for 36 volcanic centers from southern Peru. In this
chapter I explain in detail for each volcanic center the geochemical patterns, which allow me
to observe abrut geochemical variation at 16°S. This observation is compared to the northern
Chile where previous works observed geochemical variations at 19°S. This chapter indicate
me that for better understanding these geochemical variations in time and space in the Central
Volcanic Zone is also necessary to compare this results with older rocks e.g. from Jurassic to
early Tertiary volcanic arcs.
Chapter 3 has the title “Regional and temporal pattern in Meso-Cenozoic magmatic
evolution in the Central Andes (13°S to 28°S)”. For this chapter I compilied the Göttingen
Andes and selected published geochemical data of Meso-Cenozoic igneous rocks (>1500
samples) of the Central Andes, many ICPMS and Pb-Nd-Sr-isotope analysis of these samples
were performed in order to provide an adequate framework for the geochemical pattern
interpretation. The geochemical information in this database is the most complete for the
study region. With this data I could observe that significant isotopic and trace element
differences are observed for rocks of the same age but at slightly different locations. The
spatial analysis of this data is presented in the following chapter.
Chapter 4 is “Geochemical variations in time and space of lead isotopic domains in the
Central Andes (13°S -28°S): Implications of the crustal structure and metal sources”. For this
chapter I used the compiled database of igneous rocks and data of ore deposits and
sedimentary rocks under spatial analysis program at BRGM (French Geological Survey) in
collaboration of GIS Andes Project to delineate the spatial isotopic variations boundaries in
the Central Andes. This work is constrained by 3D model presented by Andres Tassara (under
review). This result offers a new knownledge of the extension, structure and composition of
the continental crust in the Central Andes.
Chapter 5 is the appendix with sample locations and explain the analitycal methods.
Chapter 6 is a list of references cited throughout the previous chapters.
11
1 Introduction
1.1 Variations in magma composition in time and space along the Central Andes (13°S-
28°S): Facts and open questions
Continued volcanic and plutonic magmatism due to subduction of the Nazca plate beneath the
South American Continent in the Central Andes (CA) is active since the Jurassic (~180 Ma)
until recent. During this time, referred to as the "Andean cycle" (Coira et al., 1982), the active
volcanic arc shifted progressively eastwards by ~200 km, thus coupling temporal with spatial
across-arc variations. Below the CA, the continental crust reaches a thickness of up to 70 km
since crustal thickening started about ~25 m.y. ago (Isack, 1988).
The potencial sources that would have contributed to create the geochemical characteristics of
the magmas in the Central Volcanic Zone (CVZ) of the Andes would be: fluids from the
subducted oceanic crust (Stern et al., 1989, 1991), asthenospheric mantle (Rogers and
Hawkesworth, 1989), lower continental crust and upper continental crust (Hildreth and
Moorbath, 1988; Wörner et al., 1988; Davidson et al., 1991), as well as melting of old
subducted oceanic crust in the shallow subduction zones (Gutscher et al., 2000). These
sources responsible for the geochemical characteristics of magmatic rocks observed in
different areas of the CVZ have faced the geocientific community since decades.
In order to have an adequate framework for this thesis, I found necessary to build a Central
Andes metadata, which contains geochemical, geochronological, geological, metallogenic and
geophysical information.
The open question is, at different sources and different processes can account for the observed
diversity of magma chemistry along and across the arc with time. Apart from assessing
contributions from different potential sources, the main interest is focused on the role of the
continental crustal composition and thickening through time and space in magma genesis.
With the metadata this work will address the following questions:
¾ Do magmatic rocks chemistry and distribution reflect changes in the dip of the subducted
Nazca plate and the thickness of the overlying lithospheric mantle and crust?
¾ In what way does crustal thickness affect the composition of magma?
¾ How is the dramatic increase of crustal thickness due to the recent uplift episode reflected
in magma geochemistry, especially the REE?
¾ Do Neogene magma composition document contributions from different basement
domains? and how is the real extension of these basements?
Further attention is given to ore deposits and their geological setting, the influence of
basement rocks and to understand the regional control on ore deposits, based on Pb isotopes
data of the mayor central Andean Neogene silver, gold, copper and tin deposits.
1.2 Working Hypothesis

The main hypothesis analysed throughout this dissertation is that igneous rocks of the central
Central Volcanic Zone are contaminated by crustal material (e.g. Wörner et al., 1988; 1992;
Davidson et al., 1990). Besides affecting Sr-Nd, and O-isotopic composition, the crustal
component has a particularly strong impact on the Pb-isotope ratios of the erupted lavas
(Tilton and Barreiro, 1980; Harmon et al., 1984; Barreiro, 1984; Mukasa, 1986; Aitcheson et
al., 1995; Wörner et al., 1992). This idea has been suggested for Cenozoic rocks, but never
formally expressed for the entire Central Andes (Stern et al., 1989, 1991; Rogers and
Hawkesworth, 1989) nor spacially and temporally explained in terms of its causes and
consequences.
During this thesis I analysed the Meso-Cenozoic igneous rocks of southern Peru and I
completed to analyse samples from northern Chile and also I analysed some new Proterozoic
12
and Paleozoic rocks for a better understanding the spatial and temporal variations of magmas
along the entire Central Andes.
1.3 Previous work
1.3.1 The Andes

The Andes is a mountain belt built along the western margin of the South American continent
with a stretching over 8000 km from Venezuela to the Tierra del Fuego, a maximum weight
of 600 km and a maximum elevation of 7000 m a.s.l. This orogenic type has produced the
highest non-collisional mountains in the world. However, when this chain is analyzed in detail
it is evident that a great variety of processes has produced it present geology. Most of its
length, the Andes consist of a magmatic arc flanked by a trench in the western and forenland
Fold-Thrust Belt and Basin in the eastern. The geological history, as well as the present
tectonic setting, is responsible for the unique geology of the Northern, Central, Southern ans
Austral Andes. The Northern Andes (10°N-3°S) are the result of Mesozoic and Cenozoic
collision of oceanic terranes, prior to the present Andean-type setting (Bosch and Rodríguez,
1992). The central Andes (3°S-33.5°S) have a long history of subduction and volcanic arc
activity (Sébrier and Soler, 1991, Isacks, 1988, Allmendinger, 1997), while the Southern
Andes (33.5°S-46.5°S) record the closing of a back-arc oceanic basin (Ramos and Kay, 1992;
Gorring et al., 1997) and the Austral Andes (46.5°S-56°S) with volcanic activity (Stern,
2004). The Andean Chain borders the Caribbean Plate in the north. The Nazca and Antartic
Plate are in the west and the Scott Plate is in the south (Fig. 1.1).
Fig.1.1. Plate-Tectonic setting of the Andes (based on the compilations of Wörner et al., 1988; Stern and Killian,
1996; Ramos 1999; Stern, 2004).
Eastward subduction of the Nazca and Antartic oceanic plate beneath the South American
Plate occurred since the Jurassic (Coira et al., 1982). The link between trench collision of
13
aseismic ridges and flat-slab segments play an important role in the volcanic activity. The
distribution of the volcanic arc has been linked to the steepening of the subduction zone and
influx of hot asthenosphere into the mantle wedge after a phase of flat-slab subduction, uplift
and no magmatism (Isack, 1988; Wörner et al., 1994; Kay et al., 1999; James and Sacks,
1999). Thorpe and Francis (1979), Stern and Killian (1996) have shown a segmentation of the
present-day active volcanism into four distinct segments with shallowly and steeply dipping
Benioff zones based on spatial seismological variations.
a) Northern Volcanic Zone (NVZ) extending from 2°S to 5°N

b) Central Volcanic Zone (CVZ) extending from 15°S to 28°S
c) Southern Volcanic Zone (SVZ) extending from 33°S to 46°S
d) Austral Volcanic Zone (AVZ) extending from 47°S to 54°S
In the NVZ, SVZ and AVZ crustal thickness is < 45 km. The flat segments with absent
volcanism and the foreland uplift correspond to the segments of Bucaramanga (north of 5°N),
Peruvian (5°-14°S) and Pampean (28°-33°S) (Fig.1.1).
1.3.2 The Central Volcanic Zone (CVZ)

The CVZ is widely developed between southern Peru, northwestern Bolivia, northern Chile
and northwestern Argentina, along the Western Cordillera, which bounds the Altiplano-Puna
plateau. Hundreds of volcanoes are widely spread along this region (Fig. 1.2). The most
striking feature of the CVZ is its extraordinary thick crust, which reaches a thickness greater
than 70 km since crustal thickening started about 25 m.y. ago (Isacks 1988). The CVZ is also
the type region of a non-accreting continental margin, as an accretionary prism is missing
opposite to the sediment-free trench along of the margin (Schweller et al., 1981). The most
important normal subduction sector is between Arequipa and northern Argentina (16°-24°S),
where the slab inclines 30° to the east. The subduction has an abrupt dip change between 14°
and 16°S in southern Peru-Bolivia, and a smooth transition between 24° and 27.5°S in
northern Argentina (Allmendigar et al., 1997). The present convergence rate is 10 cm/y
(Pardo-Casas & Molnar, 1987) 7.5 ± 0.5 cm/y in the direction N78°E (Norabuena et al., 1998;
Angermann et al., 1999) and 7.5 cm/y (Somoza, 1998).
Paleomagnetic data indicate that a 30° clockwise rotation about a vertical axis of the area
south of the bend and a 20° counterclockwise rotation north of the bend has occurred since the
Miocene. Watts et al (1995) hold regional variations in flexural rigidity responsible for the
along strike bending of the entire mountain belt.
The active volcanic front is situated 300 km east of the trench axis and 120 km above the
Benioff zone and comprises 50 recently active volcanoes.
14
Fig. 1.2. Central Andean map showing principal Geomorphogical Unites, locations of main Mesozoic and Early
Cenozoic intrusions, and Neogene volcanism.
15
1.3.2 Pre-Andean history
Proterozoic basement
Some outcrops of Proterozoic rocks are exposed along the present-day Andean margin in
southern Peru, western Bolivia and northern Chile. The major units are in the Fig. 1.3 and
Table 1.1. These contain dissimilar rock type, previous workers cited a coherent whole-rock
Pb isotopic signature throughout to define a single crustal block called Arequipa-Antofalla
basement (Tilton and Barreiro, 1980; Barreiro and Clark, 1984; Wörner et al., 1992;
Aitcheson et al., 1995; Tosdal, 1996; Wörner et al., 2000; Loewy et al., 2004).
According Wasteneys et al., 1995; Wörner et al., 2000 the oldest rocks are formed at ca. 2.0-
1.9 Ga but there has been a debate over the timing of metamorphism of these rocks. Rb/Sr and
early U/Pb studies implied granulite to amphibolite facies metamorphism between 1.9 and 1.8
Ga (Cobbing et al., 1977; Schakelton et al., 1979), but more recent U/Pb data from gneisses at
Mollendo and Cerro Uyarani indicated high-grade metamorphism at ca. 1.2-1.0 Ga
(Wasteneys et al., 1995; Wörner et al., 2000).
Tosdal (1996) and Wörner et al. (2000) propose that the Belén protoliths show a strong
sedimentary contribution from the Arequipa Massif.
According precise U/Pb geochronology, group polarity, and the sequence of adjacent
provinces Loewy et al. (2004) suggest that Arequipa-Antofalla basement was accreted, the
docking occurred during Sunsás Orogeny at ca. 1.05 Ga.
Additional Proterozoic outcrops, apparently unrelated to the Arequipa-Antofalla basement,
occur in northern and eastern of Cusco (Laubacher et al., 1984; Carlotto, 1998).
Paleozoic evolution
Paleozoic magmatism and metamorphism have been identified in the Costal Cordillera in
southern Peru, along the Eastern Cordillera and some outcrops are exposed across the western
Cordillera and Altiplano-Puna (Damm et al., 1990; Wörner et al., 2000; Lucassen et al., 2001;
Loewy et al., 2004) (Fig. 1.3 and Table 1.1).
During the Cambro-Ordovician (~500 to ~450 Ma) shallow to deep marine sediments were
deposited in a large basin in southern Peru-Bolivia and northwestern Argentina with thickness
increasing to the north (Bahlburg and Hervé, 1997). Syn-sedimentary intense volcanism
transitionally changed from tholeitic to calc-alkaline. Ordovician sedimentation and
volcanism ended with the closure of the marine basin in the Ocloyic phase (Late Ordovician
to Early Silurian) and its associated syn-cinematic granitic plutonism (Coira et al., 1982).
From Silurian-Devonian (~400 to ~350 Ma) two basins development, which were separated
by the Altiplano-Puna. During the Silurian diamictite was deposited in the pampean and
shallow-marine sandstones and Devonian shales were deposited to the east of the Eastern
Cordillera and Subandean. They were uplifted during the Ocloyic phase (~450 to ~400 Ma)
and deformed by the Chanic phase (Late Devonian to Early Carboniferous orogenesis
occurred in the Andes). Calc-alkaline plutonism occurred syn-cinematically (Coira et al.,
1982).
Carboniferous to Lower Permian (~350 to ~270 Ma) marine carbonates was deposited west
and east of the Puna-Altiplano and continental red beds east of it (Coira et al., 1982).
During Triassic occur continental sedimentary rocks in local graben structures along northern
Chile and southern Peru (Suarez and Bell, 1992; Sempere et al., 2004). Depositional
conditions changed to shallow marine in the Late Triassic/Early Jurassic. The Jurassic
sediments were deposited contemporaneously with Jurassic volcanic rocks in a north-south
16
oriented chain of basins located east of the Jurassic arc (Prinz et al., 1994; Sempere et al;
2004).
The Paleozoic evolution of the Central Andes has been and still is a subject of discussion.
Current hypotheses on the geodynamic and tectonic processes that formed the pre-Andean
basement can be summarized in the following three models (after Damm et al, 1994):
Continuous subduction, assumes a more or less uninterrupted eastward subduction beneath a
stable active continental margin with varying plate geometries since Precambrian time (e.g.
Coira et al, 1982; Pichowiak, 1994).
Exotic terrane, postulates crustal growth and consolidation during the pre-Mesozoic due to
docking of various allochthonous terranes (Pampeana, Arequipa-Antofalla, Precordillera,
Chilenia) to the stable margin of the Brazilian schield, causing deformational and
metamorphic event (e.g. Ramos et al, 1988; Dalziel & Forsythe, 1985; Astini, 1995; Tosdal,
1996; Loewy et al., 2004).
Ensialic margin, states that the formation of the western margin of South America\Gondwana
was complete by Late Proterozoic and remained autochthonous and tectonically passive until
initiation of subduction by the break-up of Pangea in the Triassic. Precambrian to Paleozoic
history is characterized by repeated closure and opening of the intracontinental basins (Damm
et al, 1990, 1994; Dalziel & Forsythe, 1985; Lucassen et al., 1996, 2001).
Table 1.1. Proterozoic and Paleozoic ages. Data are from a=Loewy et al., 2004; b=Damm et al., 1990;
c=Mpodozis et al., 1983; d=Lork and Bahlburg, 1993; e=Pacci et al., 1980; f=Cobbing et al., 1977; g=Wörner et
al., 2000; h=Lehmann, 1978; I=Tosdal, 1996.
Outcrops Rock type Protolith age (Ma) Metamorphic age (Ma)

Banded Gneiss 1819+17/-16a 1033±31a
San Juan Gneissic Tillite clast 1165a
Granite 960a
Ocoña Foliated Megacrystic granite 464+/- 4a
Banded Gneiss 1851+/- 5a 935±14a
Mollendo
Granite 468+/- 4a
Arequipa Massif Gneiss 1900f
Berenguela Ortho-paragneiss clast 1158-1080i 410-340i
San Andres Drill core 1050+/-100h 530h
Azurita Granulite 650g
Amphibole gneiss 1877+139/-131g 366±3g
Biotite-gneiss 1745+/-27g 456±4g
Amphibolite 444±14e
Belén Granodiorite 1559±21a 473±2a
Felsic dike intrudes muscovite
1866±2a 227±17a
schist
Amphibolite layers 1900 and 1000b
g
Uyarani Charnokite 2024+133/-11 1157+49/-62g
1070a and
Tonalite 450a
1254+97/-94b
Choja
1067±4a and
Orthogneiss 497±16a
1213+28/-25b
Limon Verde Granodiorite 298±1.5b
502±7c and
Cordon de Lila Granite 450+12/-11 -
434±2b
Mejillones and
Granodiorite 561+12/-14b 175±10b
Salar de Navidad
Puna Monazite 476±1d
17
Fig. 1.3. Map of Proterozoic and Paleozoic Basement outcrops (after Damm et al., 1990; Worner et al., 2000;
Lucassen et al., 2001; Loewy et al., 2004).
1.3.3 The Andean Cycle

The magmatic arc systems in the western continental margin of central South America
developed from lower Jurassic to recent times due the subduction of the Farallon-Nazca plate,
the arc progressively shifted ~150 km from a western position in the Jurassic to the present
Western Cordillera in the Oligocene time (Scheuber, 1994) (Fig. 1.2). Migration of the arc
started before ~120 Ma and is attributed to crustal erosion and variable slab dip. This period is
referred to as the Andean Cycle (Coira et al., 1982). Gradual eastward shift of magmatic
activity was mainly restricted to plutonism, whereas volcanism occasionally paused and can
be divided into periods of absent and abundant activity.
18
During the Lower Jurassic (~190 Ma), volcanism of the Precordillera shifted ~200 km
westward into the region of the present Coastal Cordillera and formed the Jurassic to Early
Cretaceous arc system. The great volumes of Jurassic igneous rocks in the Coastal Cordillera
are mantle derived (Rogers and Hawkesworth, 1989; Pichowiak et al, 1990; Lucassen and
Thirlwall, 1998).
Lower Cretaceous rift-related basins opened in a north-south corridor from northewestern
Argentina (Salfety and Marquillas, 1994) to Bolivia and Peru (Sempere, 2004) and these were
accompanied by small volumes of volcanism and alkaline intrusions (Lucassen et al., 1996b).
The break-up of Gondwana in Middle-Cretaceous times increased spreading between South
America and South Africa and enhanced the subduction rate at the western coast of the South
American continent.
During the Late Cretaceous (~80 Ma) to Early Tertiary (~60 Ma) the magmatism arc was
shifted eastward to the western foothills of the sub Western Cordillera. Several intramountain
basins, partially interconnected, also developed in northern Chile, southern Peru (Bogdanic
and Espinoza, 1994) and in northwestern Argentina were deposited in basins of Salta Rift
System (Viramonte et al., 1999), these sedimentary rocks are characterized by carbonates and
red beds.
Magmatic arc activity diminished after at ~40 Ma period during extensive red bed were
deposited in the sub Eastern Cordillera.
During the Late Oligocene and Middle Miocene renewed alkaline magmatism activity
occurred in the western border of the Altiplano-Puna. In the Late Miocene the main magmatic
arc activity increased with the onset of the higher convergence rates. The modern subalkaline
magmatic arc is 80 km broad and is located in the Western Cordillera (Coira, 1982; Wörner
1994) (Fig. 1.2 and 1.4).
1.3.4 Uplift of the Central Andes and subducting Nazca plate

Uplift of the Central Andes (Fig. 1.4) since the Eocene without collision is considered a
formidable geodynamic paradox. Uplift and the contribution of different processes are a
matter of debate. Studies assumed that crustal thickening and correlative relief increase were
caused by adding magmatic material from the mantle to the crust (Thorpe et al., 1981; Petford
et al., 1996) due to changes in the rate and angle of convergence (Pardo-Casas and Molnar,
1987; Soler and Bonhomme, 1990), absolute plate motion (Somoza, 1998), the morphology of
the subducting slab (Yañez et al., 2002) or the subduction of oceanic plateau or aseismic
ridges (Gutscher et al., 1999; Yañez et al., 2002). Recent estimates show that magma addition
would explain only a few percentages of the observed crustal thickness (Francis and
Hawkeswork, 1994; Scheuber et al., 1994; Allmendinger et al., 1997; Giese et al., 1999).
Nevertheless, magmatism has an influence on crustal rheology (Allmendinger et al., 1997)
and might locally contribute to the thickness of active arc zones (Lamb et al., 1997). Other
studies of the deformation of the upper crust and overlying sediments demonstrated that
compressional shortening of the upper plate and thrusting of the Andean chain upon the
Brazilian Shield are the driving mechanism of crustal thickening and the main phase of
tectonic shortening took place in Miocene times, which accelerated the uplift (Jordan et al.,
1983; Isacks, 1998; Sempere at al., 1990).
Estimates of shortening based on allegedly ‘‘balanced’’ cross sections and geophysical data
(Schmitz, 1994; Baby et al., 1997; McQuarrie, 2002; Müller et al., 2002) appear too small to
explain the present thickness of the whole Andean chain (Kley and Monaldi, 1998). However,
possible lower crustal flow within the central Andes may be important in the mass transfer
balance (Husson and Sempere, 2003; Hindle et al., 2005). James and Sacks (1999) developed
a model suggesting that crustal shortening is a consequence of the steepening of the slab,
caused by the heating and weakening of the crust. Models of Andean uplift consider that
19
uplift initiated ~60 m.y. ago in the Western Cordillera, developed later and slower within the
Eastern Cordillera, and accelerated in both cordilleras starting ~25–20 Ma (Gregory-
Wodzicki, 2000; Husson and Sempere, 2003). Wörner et al., 2000 concluded that crustal
shortening in the Western Cordillera is very limited, not sufficient to explain the observed
crustal thickening and the uplift is caused mainly by regional tilting between ~29 to 15 Ma
rather by crustal shortening.
Fig. 1.4. Crustal-scale section of the Andes (16°-22°S) (Compiled from Allmendinger et al., 1997; Giese, 1999).
Nazca Plate subduction

The East Pacific rise reached the North American subduction zone during Oligocene time,
being progressively subducted beneath the continental lithosphere (Pardo-Casas and Molnar,
1978; Somoza, 1998).
This event appears to have led plate boundary reorganization in the Pacific basin. Outstanding
Late Oligocene events are the breakup of the Farallon Plate into the Nazca and Cocos plate
(Pilger, 1984; Pardo-Casas and Molnar, 1978; Tebbens and Cande, 1997; Somoza, 1998). In
the western South America subduction zone the convergence during Late Oligocene-Early
Miocene (28.3 – 25.8 Ma) appears to have been dextral and being slightly oblique in Peru and
moderately oblique in Chile (Fig. 1.5b), these kinematic changes is likely related to this plate
reorganization. The Nazca-South America reconstructions at 20 Ma suggest slight obliquity at
the interpolate boundary, being dextral in Chile and sinestral in Peru (Somoza, 1998) (Fig.
1.5b).
Fig. 1.5. a) Relation between the convergence rates of the Nazca plate during the last 40 Ma along Central
Andes. b) Convergence obliquity in central Peru and in the northern Chile (After Somoza, 1998).
20
Early Cenozoic magmatism and deformation in the Central Andes were mainly concentrated
in the present fore-arc region, although evidence of uplifted zones in the back-arc region has
been reported (Kennan et al., 1995; Carlotto, 1998). The sudden change of the plate
convergence kinematic during the Late Oligocene time (28 Ma) is coeval with a generalized
eastward shift and widening of the locus of main tectonic activity, and beginning of formation
of the modern Central Andes. Outstanding events are the establishment of the Late Cenozoic
magmatic activity (Sébrier et al., 1988; Coira et al., 1982; Soler and Bohomme, 1990) and the
onset of compressional failure in wide areas of the region (Sempere et al., 1994), suggesting
that additional control, perhaps related to upper plate heterogeneities, may have existed. It
could be considered, however, that strong acceleration of convergence during Late Oligocene
time (28 Ma) was a factor contributing to mountain building in the Central Andean region.
The slowdown of convergence rate since Late Miocene time (10.8 Ma) roughly correlates
with the last deformation in the orogen. This deformation is characterized by the progressive
end of internal deformation in the Altiplano-Puna region, acceleration of plateau uplift, and
eastward jump of deformation to the Subandean and Sierras Pampeanas regions (James and
Sacks, 1999; Sempere et al., 1990). The Pliocene-Pleistocene-Holocene (4.9-0 Ma) change in
stress kinematics documented in the Puna region could be related to this slowdown of
convergence rate, as Marrett et al. (1994) hypothesized.
1.3.5 Andean volcanism

Neogene volcanism in the CVZ comprises stratovolcanoes and dome-cluster volcanoes with
predominantly andesitic and dacitic lavas (de Silva, S.L. & Francis, P.W. 1991). Basalts are
rare. Volcanic activity is frequently associated with deposition of pyroclastic flows. Large
volume ignimbrites (e.g. Oxaya-Huaylillas-Cerrillos-Acay, Lauca-Peréz-Senca-Antofalla-
Galan) with dacitic to rhyolitic compositions have been erupted from caldera complexes in a
behind the arc, mainly in Late Oligocene, Mio-Pliocene times (Wörner et al., 2000; Tosdal et
al., 1981; Coira et al., 1993). They cover areas up to 300 000 km2 and these constitutes ones
of the large ignimbrite provinces on earth (de Silva, 1989, Kay et al., 1999, Wörner et al.,
2000, Siebel et al., 2001). Pleistocene and Holocene monogenetic cones comprising spatter
cones, lava flows and proximal base surge deposits are found infrequently along the volcanic
front (Wörner et al., 1994).
Back-arc Neogene volcanic activity behind the main magmatic arc is limited to the large fault,
where rhyolitic ignimbrites to basaltic andesites flow have erupted (Soler et al., 1992; Kay et
al., 1994a; Davidson and de Silva et al., 1992; Mamani et al., 2004; Carlier et al., 2005).
Petrography
Most andesites, dacites from the arc and basalt andesites display a porphyritic texture with a
fine grained to glassy matrix. The dominant phenocryst mineral phase is plagioclase.
Hornblende is also abundant in many but not all rocks. Mafic phases like olivine,
clinopyroxene and orthopyroxene are found in basaltic andesites, biotite, sanidine and
accessory minerals such as magnetite, zircon, apatite or sphene are present in more evolved
rocks. Quartz is abundant in ignimbrite samples and sometimes also found as xenocrysts in
less evolved volcanics (Zbar, 1991).
The lavas from the back-arc are aphanitic to glassy volcanic rocks with less olivine and
clinopyroxene phenocrysts. Shoshonites series rocks locally have phlogopite phenocrysts.
Amphibole and orthopyroxene are present as microphenocrysts or resorbed phases in some
samples. Plagioclase phenocrysts are scarce or absent (Schreiber and Schwab, 1991).
21
Geochemistry
Most rocks of the main arc of the CVZ belong to the high-K calc-alkaline magma series. They
are enriched in incompatible elements compared to tholeitic low-K island arc rocks. However,
the characteristic subduction zone trace elements pattern displaying relative enrichment of
LILE (large ion lithophile elements) (Ba, K, Sr, Rb) and relative depletion of the HFSE (high
field strength elements) Nb and Ta (e.g. Pearce 1983) is maintained. Concentrations of U, Th,
Zr and K/Rb and Fe/Mg ratios are also elevated relative to island arc magmas (Wilson, 1989).
87
Sr/86Sr ratios are higher and 143Nd/144Nd ratios are lower than in rocks erupted in most intra-
oceanic setting (see Fig. 1.6). Pb isotopic compositions of CVZ are divided into three distinct
values: A low 206Pb/204Pb = 17.5 to 18.3, middle 206Pb/204Pb =18.3 to 18.7 and high
206
Pb/204Pb = 18.7 to 18.9. These three isotopically different values reflect the composition of
the underlying basement domains (Wörner et al., 1992; Aitcheson et al., 1995).
δ18O ratios of CVZ are consistently higher (6.8-14‰ SMOW) than those commonly observed
from primitive lavas from oceanic regions (5.5-6.5‰ SMOW; Harmon, R.S. & Hoefs, J.,
1984; Kyser et al., 1986).
Three geochemical group lavas in the back-arc are recognized: (1) a relative high volume
intraplate group (high K; La/Ta ratio<25), (2) an intermediate volume, high-K calc-alkaline
group (La/Ta>25), and a small-volume shoshonitic group (very high K). Shoshonitic group
lavas have generally higher incompatible elements (Th, U, Cs, LREE (light rare earth
elements)) concentrations than do intraplate and calc-alkaline group lavas. Generally these
lavas have high Sr (>0.7055) and lower Nd (εNd < -0.4) isotopic ratios (Kay et al., 1994).
Fig. 1.6. 143Nd/144Nd versus 87Sr/86Sr from rocks of the NVZ, CVZ, SVZ compared to MORB and island arcs
(from Davidson et al., 1991).
1.3.6 Magma genesis

Processes of magma genesis in the Central Andes (Fig. 1.2) have been and still are subject of
debate. Question about the structure and composition of the lithosphere, the thermal structure
beneath an arc or partitioning of elements between subducted slab derived fluids cannot be
readily resolved. The following section summarizes fundamental processes, products and
components in magma generation, which are widely agreed upon. Secondly, several
contrasting models proposed by various authors are shortly discussed.
22
General
The descending slab experiences increasing degrees of metamorphism, which leads to
dehydration and/or partial melting. Both basaltic oceanic crust and entrained sediments are
affected. The descending fluids derive from the subducted slab enrich the asthenospheric
mantle wedge in mobile LILE, lower its solidus and trigger partial melting. However, partial
melt can also be produced from the sub-continental enriched lithospheric mantle. While
passing through the continental crust, the ascending magma may stagnate and induce partial
melting of crustal material at various depths and may so be subject to contamination.
Thus, in an ocean-continent collision zone, the following components have to be considered
as potential magma sources (see also Fig. 1.7).
• The subducted oceanic plate comprising variably altered and metamorphosed
basalt with entrained pelagic or terrigenous sediments
• The sub-arc asthenospheric mantle wedge
• The sub-continental lithospheric mantle
• The continental crust
Fig 1.7. Schematic section of a subduction zones, showing the principal crustal and upper mantle components
and their interactions, to the right side a flow diagram displaying potential components involved in magma
genesis at active continental margins (after Wilson, 1989).
Models
Tectonic erosion
Stern et al., (1989, 1991) explain the elevated geochemical and isotopic characteristics of
lavas of the SVZ with incorporation of carbonate sediment and tectonically eroded crustal
material into the mantle source region by subduction. They based their conclusions on the
lack of the accretionary prism along the Chilean coast and the eastward shift of the volcanic
arc through Mesozoic to recent times. Variations in subduction geometry, type and amount of
sedimentary input and/or margin basement, volume of mantle wedge, degree of partial
melting are held responsible for geochemical and isotopic differences.
Davidson et al. (1991), however, doubt a significant contribution of subducted sediments to
arc geochemistry, since most island arcs show very restricted Sr-, Nd-, and Pb-isotope
compositions, independent of type and amount of sediments. Along the entire SVZ,
sedimentary input remains constant, as shown by Morris et al. (1990) based on 10Be isotopes.
They additionally note the relative bareness of sediments in the trench outboard the CVZ and
the relatively steep angle of subduction, which, following the Stern´s model, would contradict
the observed highly enriched geochemical characteristics of erupted lavas to be derived from
sediment subduction. Further more, no CVZ lavas display Pb isotope compositions that might
result from mixing between MORB and Nazca Plate sediments. Hildreth & Moorbath (1988)
23
argue that in SVZ in the segment (33-37.7°S) subduction geometry and sedimentary input are
constant within analytical error and are thus implausible to explain the observed geochemical
variability.
While no author questions the existence of tectonic erosion, its apparent constancy throughout
the Andean margin seems unsuitable to explain the large observed variations in erupted lavas
(Rogers & Hawkesworth, 1989).
Enriched sub-continental lithospheric mantle

Pearce (1983) postulated a significant role of enriched sub-continental lithospheric mantle in
the genesis of active continental margin magmas, aided by a subduction zone component.
Rogers & Hawkesworth (1989) explain the enriched character of CVZ lavas with partial
melting of the enriched sub-continental lithospheric mantle triggered by slab derived fluids.
They observed an eastward increase of 87Sr/86Sr ratios and incompatible element
concentrations in magma along a W-E transect at 22°S and suggested and increasing
involvement of Proterozoic lithospheric mantle beneath the Brazilian Shield. In intra-plate
settings, lithospheric mantle is generally considered to produce incompatible element enriched
magmas a low degree of partial melting. Opposed to that, however, degrees of melting are
commonly high in subduction zones (5-25%, e.g. Miller et al., 1994). In supporting their
interpretation, Rogers & Hawkesworth (1989) argue that their observed increase in 87Sr/86Sr
ratios parallels increased abundance of Sr, which is inconsistent with contamination at crustal
levels where plagioclase is stable. Fractionation of plagioclase should have lowered the Sr
content with increasing differentiation and contamination, since DplagSr is >1.
Davidson et al., (1991), however, suggest that plagioclase is not stable in deeper levels of the
70 km thick CVZ crust and Sr would behave incompatibly. Additionally, Sr is not expected to
decrease with increasing contamination by bulk mixing of mantle derived magmas and lower-
crustal melt, since plagioclase neither is a fractionating phase in mafic liquids, nor is it a
residual phase of anatectic crustal melting at high pressure.
Roger & Hawkesworth (1989) argument of eastward increasing of Ta/Sm and decreasing
Th/Ta reflecting increasing involvement of enriched sub-continental lithosphere is poorly
constrained: High Ta/Sm and low Th/Ta ratios are also estimated for crustal compositions
(Taylor & Maclennan, 1985) and could thus well be a crustal signature that introduced by
contamination.
Crustal Contamination
Leeman (1983) first describes the potencial influence of crustal structures on geochemical
composition of subduction related magmas with respect to their density driven ascent.
Beneath oceanic crust, primitive melts may rise to shallow depths of only a few kilometers
whereas at active continental margin settings, magmas stagnate at the mantle-crust boundary.
The continental crust acts as a "density filter", magmas have to differentiate toward more
acidic, less dense liquids prior to further ascent.
Hildreth & Moorbath (1988) postulate a large contribution of crustal contamination in Andean
magma genesis, since they observed a good correlation of trace element and isotope
composition with thickness of the continental crust in SVZ. The probability of contamination,
that is interaction of primitive magma with the crust, increases with the increasing crustal
thickness.
Moreover, as the depth of the stagnation zone increases, the easier the lowermost crust will
melt. In their model, ascending asthenospheric mantle-derived magmas accumulate at the
mantle-crust boundary due the decreasing density differences. Low crustal assemblages start
to melt, mix with sub-crustal magmas and slowly homogenize. In this zone of melting,
assimilation, storage and homogenization (MASH), "base-level" isotopic and chemical values
24
of erupted lavas are established. Furthermore, ascending magmas can subsequently be

modified by mid to upper-crustal AFC processes.
Strong evidence for crustal contamination is based on oxygen isotopic compositions: while
mantle δ18O-values are restricted to 6.0 ± 0.5 ‰ SMOW, crustal rock, which partially derive
from rocks that may have been exposed to long term low-temperature alteration display
higher and more variable values. CVZ lavas are characterized by significantly elevated δ18O-
values ranging from 7.3 to 14‰ SMOW (Harmon et al., 1984) suggesting a notable amount
of assimilated crust. On the other hand, crustal material could also have been incorporated
into the mantle source by subducted terrigenous sediments or crustal erosion (Stern et al.,
1990, 1991). To raise δ18O ratios from mantle values to 7-8‰ SMOW, more than 10% of
sediments need to be incorporated into the magma source in the mantle. This is unrealistic
since trace element abundances would have been extremely affected by that process (James et
al., 1982). Davidson et al. (1991) estimate de sedimentary input into the arc source of the
CVZ between 1 and 2%.
Good correlation of Pb isotope composition of erupted lavas with Pb isotope composition of
underlying basement, through which the ascending magma passed, also indicates a significant
amount of crustal assimilation.
Davidson et al. (1991) create a model for magma genesis in the CVZ, in which the most
primitive magmas found are themselves considered to be differentiates from mantle derived
magmas based on Mg#, Ni-content and isotopic composition. These "base-line" magmas are
generated in deep crustal MASH zones. During further ascent, two possible processes of
magma evolution are recognized: (1) a "closed system" trend characterized by constant
isotopic composition from basalts to dacites (e.g. Nevados de Payachata), and (2) an "open
system" trend characterized by increasing 87Sr/86Sr and 18O ratios with higher degrees of
differentiation (e.g. Cerro Galan). The first trend is assumed to result from fractional
crystallization, recharge and mixing processes in the upper crust without contamination
recorded in isotope composition. However, crustal contamination cannot be completely ruled
out, if the isotopic contrast between crust and magma is small. The second trend is considered
to result from upper crustal AFC with assimilation of high 87Sr/86Sr and 18O-rich material.
25
2 Geochemical variations in south Peruvian volcanic rocks (13°S-18°S): The role of

Abstract
The past ~29 m.y. of geologic history in the northern (13°S-18°S: southern Peru) of the
Central Volcanic Zone (CVZ) has seen increasing and decreasing magmatic production rates
and temporadic westward relocation of arc segments accompanied by abrupt chemical
changes in the magmas. These changes can be linked to underlying basement, to crustal
thickening at times of uplift in the Western Cordillera and Altiplano plateau, to mantle
lithosphere at times of frontal arc migration to the west.
The magmatic-tectonic coupling is well seen in the history: provided by new Ar-Ar ages,
major and trace element analyses, and Sr-Nd-Pb isotope analyses from 36 volcanic centers are
presented in this Chapter. Ages for the southern Peruvian centers range from Eocene (45-40
Ma) to Holocene (<0.01 Ma). Therefore, volcanic rocks of this volcanic centers record the
beginning as well as the peak of shortening and crustal thickening in the Upper Oligocene –
Lower Miocene.
The compositions of andesites erupted before and after crustal thickening are similar in terms
of major elements. However, post Miocene andesites show enrichment in trace elements (e.g.
Ba, Sr), LREE (e.g. La, Sm) and depleted in HREE (e.g. Yb). 206Pb/204Pb, 87Sr/86Sr isotopic
ratios and εNd values change abrupt at 16°S and they are similar at any given sector. Pb-
isotopes are independent of age (25-0 Ma) and Sr-Nd change with time.
Comparison of contamination indicators with age show that contamination was low in the
Tacaza arc (from 20 Ma to 10 Ma), increased sharply in the Lower Barroso arc (between 10 to
3 Ma), and remained at a high level up to the Present frontal-arc since then. These younger
volcanoes (< 3 Ma) show large ranges of Sr/Y, Sm/Yb ratios (e.g. Sara Sara, Huaynaputina).
“Adakite” magmas in southern Peru are attributed to shallow subduction of Nazca ridge. The
timing over which could this occurr show that the subduction of the Nazca rigde is not enough
shallow and hot to melt.
Depletion of Y and HREE from Miocene to Pleistocene volcanic rocks is caused by residual
garnet of crustal assimilation after crustal thickening. Any involvement of slab melts in
northern CVZ rocks to explain the “adakitic” signature is also excluded. The existing
variations in lead isotopes in samples of similar ages to the north and south in southern Peru
like the well-documeted Neogene volcanoes of the central CVZ in northern Chile, support the
notion that geochemical compositions in the magmas are controlled by the composition of the
underlying basement.
2.1 Introduction
The Central Andes (CA) are characterized by extremely thick continental crust (up to 75 km
beneath the Altiplano plateau, Yuan et al., 2002). The main crustal thickening in the CA
occurred between ~29 to 15 Ma (Wörner et al., 2000). This study was undertaken to discover
what differences exist in the nature of the arc andesites erupted before and after crustal
thickening and what these differences reveal about the processes of crustal evolution in the
magmatic arc in southern Peru. It has been established that Recent andesites from the central
and southern CVZ have compositional characteristic suggesting a high degree of crustal
contamination compared with equivalent rocks erupted in regions with thinner crust (e.g.
Hildreth and Moorbath, 1988; Davidson et al., 1991). I focus in this Chapter on andesites
erupted between ~29 Ma and Recent, a time span which brackets the events of crustal
thickening and uplift in the Western Cordillera, Altiplano plateau (Isacks, 1988, Wörner et al.,
2000, Sempere et al., 2004). Sampling was restricted to the northern Central Volcanic Zone.
The area is in southern Peru (13° to 18°S and 68.5°- 75°W). In this area an increasing number
26
of sequence-stratigraphical studies as well as mining and exploration in volcanics of the

Neogene units (Tosdal 1981; Kaneoka & Guevara, 1984; Marocco et al., 1985; Fornari et al.,
2002; Sempere et al., 2004; Acosta, 2004; Flores, 2004; INGEMMET Boletines) have
welcome the debate about names conventions. Discrepancies exist between sedimentary and
volcanology community for Neogene sediments, ignimbrites and lava flows of wich
stratigraphic positions are unknown and geochemical characterization is incomplete.
Studies of volcanic centers in the CVZ have concluded that contamination of arc magmas
occurs within the crust (e.g. Wörner et al., 1988; Davidson et al., 1991), but alternative
mechanism have been proposed such as “source contamination” by subduction erosion of the
continental margin (Stern 1991) by the influence of enriched, subcontinental mantle
lithosphere in the source region (Rogers and Hawkesworth, 1989) or by melting of old
subducted oceanic crust in the shallow subduction zones (Gutscher et al., 2000).
In this light, the purpose of this study is threefold. Firstly, I summarize the Cenozoic
stratigraphical sequence and combine Wörner’s Ar-Ar ages with other data previously
reported for Neogene volcanic unit in southern Peru between 13 and 18.2°S to invoke a
consistent naming convention. The second purpose is to present new whole rock chemical and
isotopic data for Upper Oligocene to Recent arc rocks in the northern CVZ. And the last
porpuse is to define and interpret the chemical and isotopic trends in these data in the
framework of a ~29 Ma tectonic interval. The case is made that younger magmas (< 10 Ma)
over the shallow subduction of the Nazca ridge are linked to “adakites” or slab melting
(Gutscher et al., 2000).
My goal is to better constraint the magmatic evolution in Cenozoic times and understanding
how compositional changes in arc volcanics relate to the geodynamic evolution of the
northern CVZ.
2.2 Tectonic Setting

The tectonic evolution of the Central Andes margin throughout the Tertiary has been strongly
influenced by the changes in the plate convergence rate, obliquity and steepness of the
subducting plate. In the western South American subduction zone the convergence during
Late Oligocene-Early Miocene (28.3 – 25.8 Ma) appears to have been dextral and being
slightly oblique in Peru and moderately oblique in Chile. These kinematic changes are likely
related to the plate reorganization. The Nazca-South America reconstructions at 20 Ma
suggest slight obliquity at the interpolate boundary, being sinestral in Peru and dextral in
Chile (Somoza, 1998) (Fig. 1.5b).
Geological studies of the western Andean escarpment in the southern Peru have identified a
discontinuous, punctuated history of uplift of the Western Cordillera. Sempere et al. (2004)
proposed two major stages of sedimentation based on studies of Moquegua Group sediments
and ignimbrites ages. The first stage of sedimentation was from 30 to15 Ma and the second
stage between ~10 and 2.7 Ma. Erosion has prevailed everywhere since 2.7 Ma. The age at
which the present crustal thickness was established in southern Peru is hard to define directly.
However, it is reasonable to assume that it corresponds to the cessation of the sedimentation
in the western Andean escarpment.
Kay et al. (1999) suggested that the steep REE pattern in the younger volcanics of the
southern CVZ indicate that garnet was stable in the crustal source region, and hence a thick
crust (high pressure) was established in the Pliocene.
The timing of the subduction of the Nazca ridge is important because ridge subduction would
induce flattening of the subduction angle, which in turn could lead to crustal thickening by
increasing compression in the upper plate, and to migration or broadening of the zone of arc
magmatism. Hampel (2002) proposed a model that the subduction of the northeastern end of
the Nazca Ridge began ~11.2 Ma at 11°S. The Present “flat-slab” configuration north of the
27
CVZ is at 15°S (Fig. 2.1). Another piece of evidence in support of this model may be inferred
from the correlation of the Nazca Ridge with the associated segment of low-angle subduction
and the cessation of magmatic arc activity. At present, the boundary between active and
ceased volcanism in the south and in the north, respectively, is located in the landward
continuation of the ridge, but may have gradually propagated southward due to the lateral
movement of the ridge.
Fig. 2.1. Migration history of the Nazca Ridge (from Hampel, 2002)
2.3 Description of Cenozoic units and sampled volcanic centers

The volcanoes and volcanic units before, syn and after crustal thickening of southern Peru
(Fig. 3.2) could be subdivided into different stratigraphic units using radiogenic age (e.g.
40
Ar/39Ar from Wörner unpublished data), bibliographic information, contact relations and
morphological criteria observed in the field and satellite images (Fig. 2.3).
2.3.1 Middle Eocene - Lower Oligocene, ~ 45 to ~30 Ma (Lower Moquegua Formation,

Anta Group)
The Lower Moquegua Formation outcrops along the western Andean escarpment and is
characterized by its reddish color sandstone and finer grain size sediments (mud, silt and
clay). An ignimbrite layer close to the top of the sequence (in Moquegua) has been dated at
~30 Ma by Marocco et al. (1985) and ~27 Ma (Sébrier et al., 1988). To the top of this
Formation near Cuno Cuno a marine section is exposed. These sediments were probably
deposited during a marine transgression in the Late Eocene (Sempere et al., 2004).
While the Anta Group was deposited along the NE margin of the Western Cordillera,
lithologically present alluvial and fluvial deposits; this sequence is synchronous with calc-
alkaline andesitic and dacitic volcanism ~37-38 Ma toward the base, and alkaline toward the
top ~29 Ma. This unit could be deposited in a compressive and transpressive deformation
around 44-40 Ma (Carlotto, 1998).
28
2.3.2 Upper Oligocene - Middle Miocene, ~30 to ~15 Ma (Upper Moquegua Formation,
Tacaza Group)
The Middle Moquegua Formation consists of sheet-and debris-flows of reworked volcanic
material, lacustrine sediments, and ignimbrites also deposited on the western Andean
escarpament. Within this Formation thick ignimbrites are intercalated with fine sediments
called “Huaylillas Formation” (Tosdal et al., 1981). They have been dated in Moquegua at
24.43 ± 0.17 Ma and in Puquio at 22.2 ± 0.34 Ma (Wörner unpublished data). Similar ages of
this ignimbrite are known in northern Chile (e.g. Oxaya ignimbrite, Wörner et al., 2000).
The Tacaza Group was developed along the Western Cordillera and particularly along its NE
margin. It presents an intercalation of breccias, altered lava flows, alluvial and fluvial strata
sequences and ignimbrites. Multiple intrusions of Tacaza time are observed along the Cusco-
Lagunillas-Laraqueri-Abaroa faults system. The magmatic activity of this Group took place
around 29-15 Ma (Clark et al., 1990a; Fornari et al., 2002, Mamani et al., 2004).
Some Tacaza volcanic centers of this age ~20 to ~15 Ma still form circular shield-like or
amphitheater-shaped structures. Their central depression is often erroneously interpreted as
caldera collapse. Erosion has leveled the structure to about 500-600 m above the base. Lava
Fig. 2.2. Sample locations in southern Peru: Active volcanic (0 - 0.8 Ma) 7) Sara Sara, 12) Andagua, 13)
Huambo, 45) Iquipi, 21) Chivay, 28) Nicholson, 15) Sabancaya, 27) Chachani, 29) El Misti, 31) Ubinas, 33)
Huaynaputina, 34) Ticsani, 35) Tutupaca, 36) Yucamane, 37) Casiri, 38) Titiri, 39) Kere, 40) Purupurine. Upper
Barrosos (0.8 - 3 Ma) 9) Firura, 10) Antapuna, 11) Coropuna, 14) Hualca Hualca, 26) Paquetane. Lower
Barrosos (3- ~10 Ma) 8) Yarihuato, 20) Tuti, 22) Huarancante, 17) Morane, 23) Ananto, 24) Hualto, 25)
Huacullani, 30) Salinas, 32) Pichu Pichu, 41) Tarata. Tacaza (15- ~25 Ma) 5) Puquio, 6) Cora Cora, 18)
Condoroma, 19) Colca, 43) Sta. Lucia, 44) Cotahuasi. Backarc (< 1 Ma) 1) Pisaac, 2) Rumicolca, 3)
Oroscocha, 4) Quinsachata. (15 - 6 Ma) 46) Lago Titicaca, 47) Chignaya. 16) Anta (~40 Ma). Proterozoic and
Paleozoic basements 42) Huacano, 49) Arequipa, 48) Pescadores. White line is the distance used in the
chemical diagrams.
29
flows consist of aphanitic to porphyritic andesite. Many of the lavas are altered. However,
samples collected for chemical analysis from Puquio, Cora Cora, Condoroma volcanoes were
unaltered.
2.3.3 Upper Miocene – Lower Pliocene, ~10 to ~3 Ma (Lower Barroso Formation)

Lower Barroso Formation developed along the western Cordillera and preserves the eroded
stratovolcanoes structures with slightly erosion and hydrothermal alteration in the interior of
the stratocones. The eroded lavas from such volcanoes intercalated with mud are observed
toward the west of Lower Barrosos volcanoes (e.g. within Chili, Challahuaya valley). Tosdal
et al. (1981) reported a plagioclase age for Cerro Barroso of 5.3 ± 0.3 Ma and 7.0 ± 0.4 Ma
plagioclase age, 3.3 ± 0.1 Ma whole rock age for andesites near Cerro Barroso. Bellón and
Lefévre (1976) reported ages of 4.45 and 4.10 Ma for stratovolcanoes northeast of Arequipa.
Klinck et al. (1986) obtained ages: from Hualto volcano (6.1 ± 0.6 Ma), Ananto (9.5 ± 0.6
Ma), Huarancante (6.7 ± 0.7 Ma) and Tuti (5.3 ± 0.7 Ma). Kaneoka & Guevara (1984)
reported an age of 7 Ma for Salinas volcano and 6 Ma for volcanoes near Lake Titicaca and
between 6 and 3 Ma for volcanoes around Tarata.
Stratovolcanoes of Lower Barroso arc are comprised of flat shieldvolcanoes. Exposures
suggest that most of the volcanoes are made up of andesites and dacites. They represent major
volumes. Based on their state of erosion they fall into this group of stratovolcanoes and
represent its largest examples of the Barroso arc. The samples collected correspond to
Yarihuato, Huarancante, Morane, Ananto, Hualto, Huacullani, Pichu Pichu, Salinas, Tarata
and many other deeply glaciated volcanoes.
Mio-Pliocene ignimbrite centers are not observed; pumices were taken from the ignimbrites
units of the Lower Barroso Formation, these samples were collected in Condoroma, Pampa
Cañahuas, Pausa, Caraveli, Cotahuasi, Ocoña, Yura, Sumbay, Chachani and Aguada Blanca
(see on CD).
2.3.4 Pliocene, ~3 to ~2 Ma (Sencca Formation)

The Sencca Formation is deposited between the Western Cordillera and NW of the Altiplano
plateau, consist of lacustrine sediments and ignimbrite flows filling the incised valleys (e.g.,
Sencca, Yauri basins). We observed the Sencca ignimbrite on the top of Lower Barroso lavas
near the Sencca village. This ignimbrite is correlated to the 2.72 Ma Lauca-Peréz ignimbrites
of northern Chile (Wörner et al., 2000). For this study pumices of similar age were taken in
Maure, Moquegua, Chuquibamba, Salamanca and Caraveli.
2.3.5 Upper Pliocene to Pleistocene, 3 to 0.8 Ma (Upper Barroso Formation)

The Upper Barroso Formation represents recently active volcanoes along the Western
Cordillera. Some of its stratocones (e.g. Coropuna, Hualca Hualca and Chachani) have been
active up to the Pleistocene and Holocene. Kaneoka & Guevara (1984) reported ages for
Antapuna volcano of 1.19 and 1.2 Ma. Klinck et al. (1986) obtained ages from Hualca Hualca
lavas (1.1 Ma) and Chivay lavas (0.9 Ma). The typical evolution is characterized by
continuous eruption of largely similar andesite or rhyo-dacite lava composition. Other
volcanoes with such characteristic are Firura, Paquetane. Our sampling covers lavas from all
volcanoes mentioned. The lavas are generally fresh andesites.
2.3.6 Holocene Volcanoes < 0.01 Ma

The Recent volcanic arc groups the active volcanoes. They are ~230 km to the east of the
Peru-Chile trench and ~150 km above the Benioff-Wadati plane (Cahill and Isacks 1992).
Most of the youngest and better known volcanoes: active with historical eruption <0.5 Ma
(Sabancaya, El Misti, Ubinas, Huaynaputina, Ticsani, Tutupaca, and Yucamane); and
30
dormant < 0.8 Ma (Sara Sara, Ampato, Casiri, Purupurine, Titire, and Kere) are stratocones
with symmetrical shape and central vent. The lavas from these volcanoes consist of
hornblende-plagioclase andesites to dacites with a generally unaltered, glassy groundmass.
The long lived (< 3 Ma) Chachani dormant volcano is a stratovolcano complex of dacitic
domes, andesitic lavas and with pyroclastic surface.
Other Pleistocene to Recent volcanoes are:
Shoshonites
Glassy volcanic rocks (< 0.6 Ma) erupted from fissures associated with fault related to a
change in the regional stress system in the back-arc, such as Quinsachata, Oroscocha,
Rumicolca, and Pisaac volcanoes.
Monogenetic
Andesitic lavas (< 5000 yr B.P) erupted from small cinder cones, which produced steep-sided
cones and sometimes ribbonlike flows (de Silva and Francis, 1991), such as Andaguas (Puca
Mauras, Tischo, Jenchanya, Ninamama, Chilcayoc), Huambo, Cerro Nicholson, Chivay,
Iquipi and many others.
Fig. 2.3. Schematic profiles of Cenozoic stratigraphic units in southern Peru. After Tosdal et al. (1981), Kaneoka
& Guevara (1984) and Sempere et al. (2004). In italic are the Moquegua stratigraphic units presented by
Sempere et al. (2004).
2.4 Results
2.4.1 Geochemistry
The geochemical data set on which this discussion is based contains nearly 625 analyses from
36 volcanic centers and ignimbrites. Representative data from each center and the full data set
is available on CD (opening: database and samples location).
31
Major elements
All Neogene samples of this study in southern Peru plot in the calc-alkaline field of the AFM-
diagram (Fig. 2.4). The vast majority of samples form a linear array displaying a typical calc-
alkaline trend. However, this trend has two groups: lavas with more FeOtot and MgO, and
ignimbrites with higher content in alkalis.
Fig. 2.4. AFM-diagram after Rickwood (1989) of sourthern Peru data
Fig. 2.5. The subdivision of subalkalic rocks diagram of volcanic rocks from southern Peru. After Rickwood et
al. (1989).
The samples plot mainly in the “high K” field of the K2O vs. SiO2 diagram (Fig. 2.5).
Exceptions are Eocene lavas, which belong to the shoshonitic series. SAR-00-05 (Sara Sara),
BAR-00-36 (Lower Barroso- near rio Tambo), YAH-00-17 (Yarihuato) fall into the “medium
K” fields. A few Eocene samples plot in the shoshonitic field. The most basaltic lavas are
rather rare and have SiO2 contents down to 54%. The Miocene lavas lack samples with silica
32
content higher than ~66 wt%. The ignimbrites display a very restricted silica range from 72-
77 wt% SiO2 and highly variable potassium content at a constant SiO2 content.
Total alkalis versus silica are plotted in the TAS diagram (Fig. 2.6). The main data define a
more or less linear array from basalt to rhyolite. In contrast to the K2O vs. SiO2 diagram, the
ignimbrites do not deviate from this trend due to its lesser Na2O content at higher silica
content.
Fig. 2.6. TAS diagram after Wilson (1989) of southern Peru data.
There is little variation in major element composition with age. Figure 2.7 shows that
Holocene, Pleistocene and Mio-Pliocene volcanoes are more variable in major elements
contents and display the largest ranges than Miocene and Eocene volcanoes. No clear trend of
TiO2 and MgO are observed between the Neogene volcanoes.
Trace elements
Spatial and temporal variations of selected trace elements are shown in Fig. 2.8. Elements
have been analyzed by XRF (e.g. Rb, Sr, Ba, Y) and ICPMS (e.g. Nb and Zr).
Concentration of the LIL elements Ba and Sr clearly increase with younger lavas. Contents of
Zr and Nb do not change in the younger lavas. Y concentrations are highest in Miocene and
Mio-Pliocene lavas and the minimum Y values decrease from Mio-Pliocene to present.
33
Fig. 2.7. Major elements of samples from southern Peru versus distance between volcanoes (see distance line in
the Fig. 2.2). The numbers represent the volcanoes names (see Fig. 2.2).
34
Fig. 2.8. Trace elements of samples from southern Peru versus distance between volcanoes.
Spiderdiagrams
Primitive mantle normalized spiderdiagrams (primitive mantle values from McDonough et al.,
1992) of trace elements plus potassium and titanium for Neogene volcanoes are shown in Fig.
2.9.
Lavas
The lavas display a relatively uniform typical arc signature characterized by strong
enrichment of Rb, Ba, K and Th combined with a marked depletion at Nb and Ta (e.g.,
Pearce, 1983). U is also enriched, but shows highly variable concentrations. Samples from the
Plio-Pleistocene to Holocene volcanoes display positive Sr anomalies. The Sr troughs occur
within the Miocene and Eocene lavas (Fig. 2.9).
Ignimbrites
Similar to the lavas, both the Mio-Pliocene and Miocene ignimbrites display pronounced Nb-
Ta depletion (see Fig. 2.10), probably reflecting the large amount of crustal material in their
evolution. Negative Sr anomalies are strongly developed, especially in the younger
ignimbrites. The ignimbrites also have marked Ba and Ti troughs.
35
Shoshonites
Shoshonite backarc samples have positive Ba anomalies compared to the Pleistocene and
Holocene lavas. Ta is not depleted as in the lavas from the arc. They are higher in La, Ce
anomalies and lower in Y than the lavas from the arc.
Fig. 2.9. Primitive mantle normalized concentrations of trace elements plus K and Ti of each volcano.
Normalization factors are from McDonough et al. (1992).
36
Fig. 2.10. Primitive mantle normalized concentrations of trace elements plos K and Ti of Neogene ignimbrites.
Normalization factors are from McDonough et al. (1992).
Rare earth elements (REE)

The overall shape of chondrite normalized (McDonough and Sun, 1995) REE element pattern
of each volcano with respect to time and space will be described first in this section. Then,
element ratios will be evaluated quantitatively.
Description of REE pattern

The slope of the REE patterns become steeper with decreasing age of the volcanoes (Fig.
2.10), as the HREE become more depleted and the concentrations of LREE are higher and
more variable. No negative Ce anomaly or systematic trend in size and no Eu anomalies is
observed in the lavas. A conspicuous feature of most samples is a flat HREE pattern (except
37
for Monogenetic, Huaynaputina and Ticsani volcanoes). All ignimbrites samples show
marked negative Eu-anomalies and depletion in HREE (Fig. 2.11).
Fig. 2.11. Chondrite nomalized REE-pattern of lavas. Nomalized constant are taken from McDonough and Sun
(1995). All samples have been analyzed by ICPMS.
38
Eocene and Miocene volcanoes

The Eocene and Miocene lavas display a relatively flat pattern with enrichment of LREE and
a negative Eu-anomaly. They have the largest concentration of HREE (e.g. Ho, Er, Tm, Y) of
all samples studied and the HREE also are parallel.
Mio-Pliocene volcanoes
The pattern of these volcanoes is very uniform, which display elevated pattern with LREE
enrichments.
Lavas of Yarihuato, Cotahuasi, Chuquibamba, Tuti and Morane volcanoes have fixed HREE
and variable LREE. Samples of Hualca Hualca, Paquetane, Hualto, Ananto, Salinas and Pichu
Pichu volcanoes have parallel REE.
Pliocene-Pleistocene-Holocene volcanoes
The Pliocene to Holocene lavas show enrichment of LREE and depleted HREE
concentrations compared to all samples.
Samples of Sara Sara, Antapuna, Coropuna, Chivay volcanoes have parallel LREE and fixed
HREE. Lavas of Yucamane, Tutupaca, Titire, Kere, Purupuruni volcanoes show variable
HREE. Ticsani and Huaynaputina lavas have parallel LREE, while Chachani, El Misti and
Ubinas lavas have variable MREE and HREE.
Mafic volcanoes (monogenetic)

Mafic lavas show enrichment of LREE (principally La, Ce, Pr, Nd) and depleted HREE (Er,
Tm, Yb, Lu) concentrations. They have fixed LREE and variable HREE.
Ignimbrites
Samples of Miocene and Mio-Pliocene ignimbrites show a trough at Dy-Er and elevated Yb
and Lu concentrations, forming a dish-shaped pattern. Most ignimbrites of all ages display
marked negative Eu anomalies.
Shoshonites
As seen in the Figure 2.12, REE pattern of shoshonites show LREE enrichment, variable
LREE levels, small Eu anomalies. HREE elements are like the Mio-Pliocene lavas.
39
Fig. 2.12. Chondrite nomalized REE-pattern of Neogene ignimbrites and Pleistocene shoshonites. Nomalized
constant are taken from McDonough and Sun (1995). All samples have been analized by ICPMS.
2.4.1.1 Temporal variations of element ratios

In order to quantify the slope of REE pattern through time, selected chondrite normalized
REE ratios are plotted versus age in Fig. 2.13. Along arc, maximum La/Yb ratios increase
from Eocene to Holocene rocks. Also, ratios become more variable with younger ages.
Sm/Yb ratios and La/Yb ratios resemble the same pattern in time. They show a pronounced
increase with decreasing time.
La/Sm ratios have a base-line constant (La/Sm = 4).The lavas display a trend that increase
slightly with decreasing age. The ignimbrites exhibit high La/Sm ratios than the lavas.
40
Fig. 2.13. REE ratios of lavas and ignimbrites of southern Peru.
2.4.1.2 Isotopes
Sr and Nd isotopes
Sr and Nd isotopes in igneous rocks can be used to determine the presence or absence of
recycled continental crust. They are a powerful tool for constraining the age of magmatic
sources and distinguishing between mantle, upper-crustal and lower-crustal sources
(Hawkesworth and van Calsteren, 1984).
As shown in Fig. 2.14, variable 87Sr/86Sr ratios and εNd values are observed in Peruvian
volcanoes (PV). The Eocene (Anta lavas) and Miocene lavas from Cora Cora, Colca
volcanoes have the highest εNd and lowest 87Sr/86Sr. They overlap with data from the Southern
Volcanic Zone (SVZ) of Neogene volcanoes. Nd-isotopes ratios decrease and Sr-isotopes
ratios increase with time for volcanic centers between 14°S and 16°S (PV-A). High 87Sr/86Sr
ratios and low εNd ratios are observed between 16°S and 18°S (PV-B) with volcanoes toward
the northeast of Arequipa city (El Misti, Salinas, Chachani, Pichu Pichu, Paquetane) having
the highest Sr-isotopes and lowest Nd-isotopes ratios.
41
the Peruvian ignimbrites (PIG-A) around Puquio, Cotahuasi, Salamanca, Chuquibamba areas
from 14°S to 16°S have high εNd values. Low Nd-isotope ratios are observed in PIG-B across
the Moquegua region between 16.5°S and 18°S. The lowest εNd values have been analyzed for
the Arequipa ignimbrites (Fig. 2.14).
Monogenetic volcanoes (e.g. Andagua and Huambo), and Shoshonite volcanoes (Rumicolca,
Oroscocha, Quinzachata, Pisaq) have intermediate Sr-Nd isotopes ratios compared to all
samples from southern Peru.
Fig. 2.14. 87Sr/86Sr versus εNd of samples from southern Peru compared to Southern Volcanic Zone (SVZ) and
Middle Oceanic Ridge Basalt (MORB). BSE is bulk silicate earth. Data of SVZ are taken from Kay et al. (2005).
Lines mark the Mantle Array. PIG-A Peruvian Ignimbrites and PV-A Peruvian Volcanoes between 14°S and
16°S. PV-B Peruvian Volcanoes and PIG-B-Peruvian Ignimbrites between 16°S and 18°S. Number of the
symbols in Fig. 2.2.
Spatial and Temporal variations of Sr-Nd isotopes

87
Sr/86Sr and 143Nd/144Nd analysis cover uniformly the area along the arc of southern Peru (see
database).
In Fig. 2.15 87Sr/86Sr ratios divides the arc into different groups. Further can be distinguished
between lavas and ignimbrites. The northernmost groups of lavas (0 and 350 km of arc
measured from NW tip) has a Sr-isotopes = 0.7045-0.707. Further south from 300 to 600 km
arc distance, a deacrease in 87Sr/86Sr ratios 0.708-0.7065 to 0.707-0.706 is observed.
The ignimbrites have higher 87Sr/86Sr ratios (up to 0.71) than the lavas and the lowest Sr-
isotope ratio (0.7055) is observed between 0 and 300 km.
The 143Nd/144Nd ratios display two groups independent of rocks type along the arc. The
northern group (from 0 to 350 km) has higher ratios (0.5124 to 0.5127), while the second
cluster further south (200 to 600 km) shows lower ratios (0.5124-0.5120). Monogenetic
volcanoes lie in both groups and shoshonites display small differences for Nd-isotope ratios.
If ploted against age (Fig. 2.16) 87Sr/86Sr ratios increase and 143Nd/144Nd ratios decrease over
time independent of spatial affiliation. The strongest variations of Sr-Nd isotopes ratios are
42
observed in Mio-Pliocene lavas (~7 Ma) and in Miocene ignimbrites (~15 Ma). Lavas
between 10 and 0 Ma display more variable Sr-Nd isotope ratios than the older lavas (15 -30
Ma).
87
Fig. 2.15. Sr/86Sr and 143Nd/144Nd ratios versus distance. Legend as in Fig. 3.14.
Fig. 2.16. 87Sr/86Sr and 143Nd/144Nd ratios versus Age.

43
Lead isotopes
Lead isotopes data fall into two groups independent of age and rocks type: the most
radiogenic samples range in 206Pb/204Pb ratios ranging from 18.5 – 19. More unradiogenic
samples show ratios of 17.8 to 18.5 (see Fig. 2.17, 2.18). 207Pb/204Pb and 208Pb/204Pb ratios
increase slightly with increasing 206Pb/204Pb.
Isotopic ratios observed in some volcanoes (e.g. El Misti, Huaynaputina, Huambo and
Andaguas) have unusually low 207Pb/204Pb ratios ranging from 15.50 to 15.55.
The samples of southern Peru in the 207Pb/204Pb versus 206Pb/204Pb diagram lie between the
evolutions lines of an upper crustal lead and orogenic belt lead signature. In the 208Pb/204Pb
versus 206Pb/204Pb diagram all samples plot above the orogenic belt lead line of Zartman and
Haines (1988).
Fig. 2.17. Lead isotopic compositions of the Neogene rocks in southern Peru. Zartman & Heines (1988) lead
evolution model: M. upper mantle lead; O. orogenic belt lead; U. upper crustal lead. The ticks of each curve are
at 100 Ma interval.
Spatial and Temporal variations of Pb-isotopes

Samples with higher 206Pb/204Pb ratios (18.5 to 19) are located exclusively north of Huambo
volcanoes. The low 206Pb/204Pb ratios (17.8 to 18.5) are located to the south of Colca canyon.
The lowest 206Pb/204Pb (17.5-17.8) ratios have El Misti, Salinas and Paquetane volcanoes.
Transitions between these groups are abrupt as seen in Figure 2.19.
Ignimbrites have the same Pb-isotope signature as the other volcanic products of their area.
Neogene ignimbrites erupted around Moquegua and Arequipa regions have 206Pb/204Pb ratios
similar to Chachani, Pichu Pichu, Ubinas, Huaynaputina, Yucamane, Casiri, Tacora
volcanoes. Ignimbrites erupted around Chuquibamba, Cotahuasi, Puquio, Caravely, Cuno
44
Cuno, Pausa, Yauri have 206Pb/204Pb ratios similar to Coropuna, Antapuna, Cora-Cora, Sara
Sara and Condorama volcanoes.
Fig. 2.18. 206Pb/204Pb versus distance. Numbers represent the volcanoes positions (see Fig.2.2). Legend as in Fig.
2.17.
The 207Pb/204Pb ratios change slightly with decreasing age of volcanic rocks. The unusually
lowest ratios observed for Holocene lavas correspond to the El Misti, Huaynaputina Andaguas
and Huambo volcanoes. The 206Pb/204Pb ratios in both groups (radiogenic and unradiogenic)
changed also with decreasing age of volcanic rocks (Fig. 2.19).
207
Fig. 2.19. Pb/204Pb and 206Pb/204Pb versus Age. Legend as in Fig. 3.17.
45
2.5 Discussion
This study has documented systematic spatial and temporal differences in trace element and
isotope signature among lavas and ignimbrites rocks erupted at different ages and at different
regions in the northern CVZ.
The compositional differences between MioPliocene-Recent and Eocene-Miocene lavas are
observed in incompatible elements ratios (e.g. Sr/Yb, Sm/Yb, La/Yb). Compositional
differences between Neogene lavas and ignimbrites exist for incompatible elements ratios of
La/Sr, Sr/Y, La/Sm. Higher Sr/Y and Sm/Y or Sm/Yb between Miocene (Lower Barroso),
Mio-Pliocene volcanic (Upper Barroso) and Frontal Arc volcanics in southern Peru is mostly
controlled low Y and Yb in the younger rocks (<10 Ma) principally the volcanoes between 14
and 16°S. Thus rocks in southern Peru plot in the calc-alkaline field (Fig. 2.4), have a gradual
slope to higher REE pattern (La/Yb = 70, Fig. 2.13) that reflect both light REE enrichment
and heavy REE depletion (Fig. 2.11), small Eu anomalies and high La/Ta ratios. Andesites
and dacites (SiO2 = 57-67%) are distinguished by their high La/Ta ratios (to 160, Fig. 2.20),
the steep REE pattern (La/Yb to 70) and high Sr/Y ratios (to 100, Fig. 2.21).
Other notable features are the less negative 143Nd/144Nd values (0.5126-0.5124), high 87Sr/86Sr
values (0.705-0.7065) and high 206Pb/204Pb (>18.5) ratios in the Upper Barroso and Recent
volcanoes (e.g. Sara Sara, Yariahuato, Andaguas, Coropuna, Firura, Ananto) (Fig. 2.2, 2.14
and 2.18).
Magmas with chemical features like those in the andesites, dacites and young volcanics are
now commonly called adakites (Defant and Drummond, 1990; Gutscher et al., 2000; Beate et
al., 2001). However, the setting for the formation of adakites is related to the melting of the
subducted hot and young oceanic crust, it is only during the early stages of flat subduction
that the upper part and leading edge of the slab can melt, and prolonged flat subduction will
cool both lithospheres and impede partial melting. In this sense Gutscher et al. (2000)
presented a flat subduction slab melting model and the emphasis of their model is not on the
initial temperature of the slab, but on its buoyancy and thus on the time spent at a relatively
shallow (~80 km) depth, allowing the slab to heat up. Therefore, oceanic crust as old as 50
Ma can melt if it is given sufficient time to warm up.
In the discussion below, the regional setting, chemical an isotopic characteristic of the
northern CVZ are compared to, and evaluated relative to, similar north and south trend in
central CVZ (North Chile, Wörner et al., 1988). The chemical parallels seen in the northern
and central CVZ and regional geologic observations, are used to argue that Miocene to
Holocene trends in southern Peru magmas and northern Chile magmas have similar origins.
The case is made that crustal contamination is needed to explain these trends and not slab
melting. A simple geochemical parameter such as Sr/Y (Fig. 2.21) is clearly insufficient to
make the case for slab melting (Kay, 2002; Garrison & Davidson, 2003).
46
Fig. 2.20. SiO2 versus La/Ta, La/Yb and Sr/Y.
Fig. 2.21. Y versus Sr/Y diagram showing the evolution of the northern CVZ in southern Peru before, during and
after crustal thickening. Adakite and calc-alkaline fields after Defant, M.J., & Drummond, M.S. (1990).
2.5.1 Chemical isotopic composition of magmatic rocks in southern Peru and northern
Chile
The Temporal trends in northern CVZ (13-18°S) magmas discussed above are highlighted by
calc-alkaline trace elements compositional have steepening REE patterns and isotopes varies
47
from north to south. On the basis of published data for central CVZ (18-22°S) magmas (Fig.
2.22, Wörner et al., 1988; 1992 and references therein), these trends are generally similar to
those from south to north along the northern CVZ as shown by the following observations: (1)
Neogene volcanic rocks between 16-18°S are like volcanics from 18-19.3°S; (2) samples
from 14 to 16°S are like samples between 20.5 and 22°S; (3) in the transition zone in northern
CVZ there is not a volcanic gap like in the transitional zone of the central CVZ.
Fig.2.22. Morphological map of the area central CVZ between 18 and 22°S, with Plio-Pleistoceno and Miocene
volcanic centers, and outcrop areas of intrusive rocks after Wörner et al., 1988.
These broad regional comparisons of central CVZ and northern CVZ magmas are intended to
emphasize their dominant characteristics in the long tradition of studies along the CVZ
(Wörner et al., 1988, 1992; Davidson et al., 1991). On a finer scale, the composition of the
magmas from some central CVZ centers have been show to be quite diverse, as documented
by detailed studies of the Nevados de Payachata ( Pomerape and Paricota volcanoes, Wörner
et al., 1988, Davidson et al., 1990).
The chemical variations in the northern CVZ magmas, like those along the central CVZ, can
be attributed to first order changes in the chemistry of the magma source and the residual
mineral assemblage that equilibrated with the magmas. Other modeling studies in the Andes
(e.g. Kay et al., 1991, 1999; Hildreth and Moorbath, 1988; Davidson et al., 1991; Haschke et
al., 2002) and elsewhere (e.g. Aleutians; Kay 1978) show that steep REE pattern that are with
high Sr and high Na concentration can be explain by the last equilibration of these magmas
with high-pressure plagioclase-poor, garnet-bearing residual mineral assemblages.
Changes in 87Sr/86Sr ratios in the northern CVZ samples (Fig. 2.14) also show a Mio-Pliocene
to Holocene progression similar to south and north trends along the central CVZ (Fig. 2.23;
data summaries from Wörner et al., 1988, 1992, unpublished data and 30 new samples from
this study).
48
Fig. 2.23. Diagram of εNd vs. 87Sr/86Sr for northern CVZ (14-18°S) and central CVZ (18-22°S). HV-Holocene
volcanoes, PPV-Pliocene Pleistocene volcanoes, MPV-Mio-Pliocene volcanoes, MV-Miocene volcanoes, EV-
Eocene volcanoes.
Especially, εNd values and 87Sr/86Sr ratios in: 1) Miocene, Mio-Pliocene volcanoes (14-16°S),
Mio-Pliocene volcanoes (19.5-22°S) samples either overlap or are depleted (higher εNd and
lower 87Sr/86Sr) than those from Miocene, Mio-Pliocene volcanoes (16-18°S), Mio-Pliocene
volcanoes (18-19.3°S); 2) Plio-Pleistocene to Holocene volcanoes samples from (16-18°S)
overlap those from central CVZ centers (18-19.3°S); and 3) Plio-Pleistocene to Holocene
volcanoes samples from (14-16°S) overlap those from central CVZ (19.3-22 °S).
As with changes along the northern CVZ (14-16°S) and (16-18°S), distinct shifts to lower εNd
values and higher 87Sr/86Sr ratios between Eocene-Miocene samples and Mio-Pliocene-
Pleistocene-Holocene volcanics after ca. 10 Ma (Fig. 2.24) are nearly independent of SiO2,
Nd, Sr contents (Fig. 2.24, 2.25). These shifts thus reflect differences in compositions and
amount of crustal components. To maintain the narrow range in isotopic values seen across
the broad range of SiO2, contents in each period and zone (Fig. 2.24), the evolution of the
Tacaza magmas (~ 20-15 Ma, from 14-16°S) can be involve only minor subsequent crustal
contamination or, less likely, only contaminants with low Sr and Nd contents and/or with
isotopic ratios like those of the Anta lavas (Fig. 2.23).
49
Fig. 2.24. Diagram of εNd vs. SiO2 for northern CVZ samples (number of samples as in Fig. 3.2).
Figure 2.25 show that the Upper Barroso and Holocene lavas of both zones (14-16°S and 16-
18°S) are enriched in incompatible trace elements (e.g Ba, Sr). This reflects the response of
increasing contamination in a thick continental crust (70 km), so that contaminated parent
magmas may occur at depths between 50 and 70 km (lower crust) and at pressure between 15
and 20 kbar where plagioclase is unstable and Sr therefore behaves incompatible. Contrary Sr
behaves compatible in Peruvian Ignimbrites indicating that contaminated parent magmas
occur where plagioclase is stable, followed by upper crustal differentiation (fractionation
and/or AFC).
Fig. 2.25. Diagram of 87Sr/86Sr versus Sr for northern CVZ samples (number of samples as in Fig. 3.2).
50
In contrast to Nd, and Sr isotope ratios in each zone, there is not obvious increase of
206
Pb/204Pb ratios with decreasing ages. The abrupt Pb isotope changes are at 16°S in the
northern CVZ and at 20.3 °S in the central CVZ (Fig. 2.26). Especifically, the samples located
between 16 and 18°S in southern Peru resemble the samples from 18 to 19.3°S in northern
Chile. These samples have the lowest ratios, whereas volcanic samples from 14 to 16°S are
like the volcanic samples between 20.3 and 22°S.
Wörner et al. (1992) and Aitcheson et al. (1995) argued that Pb isotope ratios in central CVZ
(18 - 22°S) centers strongly reflect contamination from the underlying basement, as ratios of
volcanoes from (14-16°S) and (16-18°S) in southern Peru are like those in basements
samples, or co-existing ignimbrites in the close vivinity (Fig. 2.27).
A local crustal contamination is consistent with mass-balance consideration, as Pb is strongly
concentrated in feldspar-rich crust compared to peridotite mantle.
End-member model to explain Nd-Sr isotope variations along the northern CVZ agree well
with models for trace elements. In the northern CVZ the MASH model can be applied
(Hildreth and Moorbath, 1988). This model calls upon isotopic homogenization of mantle and
crustal components at the base of the thicker crust, more radiogenic crust in the north (14-
16°S) and unradiogenic crust to the south (16-18°S) like the central CVZ unradiogenic crust
to the north (18-19°S) and more radiogenic to the south (20-22°S, Wörner et al., 1992).
Fig. 2.26. Diagram of 206Pb/204Pb versus Latitude S for northern central volcanic zone (NCVZ) and central
volcanic zone (CCVZ) samples. HV-Holocene volcanoes, PPV-Pliocene Pleistocene volcanoes, MPV-Mio-
Pliocene volcanoes, MV-Miocene volcanoes, EV-Eocene volcanoes, PIG-Peruvian ignimbrites.
51
Fig. 2.27. Diagram of 207Pb/204Pb versus 206Pb/204Pb for northern CVZ volcanics. Number of samples as in Fig.
3.2. The upper crust (UC), orogen (OR) and mantle (M) evolution curves are from Zartman and Doe (1981).
The slab melting model of Gutscher et al. (2000) that calls for slab melting by flat slab
subduction (e.g. Nazca ridge) is based on the geochemistry of erupted lavas (Fig. 2.21).
Counter arguments against the slab melting model are based on geophysical, geological and
detailed geochemical data, which show that regions of flat subduction in the Andes
correspond to lack of volcanism (Cahill and Isacks, 1992). There is not an equally satisfying
correlation between flat slab location and subducting seamount. Geometry of the slab angle
shows that the dip of the Nazca ridge is steeper than Juan Fernandez and Carnegie ridges (Fig.
2.28, Gutscher et al., 2002).
Fig. 2.28. 3D perspective image of the subducted Nazca plate (see Carnegie, Inka, Nazca, Iquique and Juan
Fernández ridges) viewed from the southeast (from Gutscher, 2002).
Isotope and trace elements variations in volcanoes above the southern edge of the Nazca ridge
are not from slab melting. These variations, which are reflected principally in Pb isotopes
52
show an assimilation of the underlying basement radiogenic to the north (14-16°S) and
unradiogenic to the south (16-18°S, Fig. 2.27). Variation in the REEs, is due to the shift in the
combined source and fractionating residual minerals assemblage from low-pressure
anhydrous pyroxene-feldspar-dominant assemblage to a medium-pressure hydrous
amphibole-feldspar-dominant assemblage to a finally higher-pressure garnet-bearing, feldspar
poor assemblage (Kay et al., 1994). The thickness of the crust in southern Peru is > 70 km
(Yuan et al., 2002) which correspond to pressures of ~21 kbar at the base of the crust, where
garnet is stable. Plagioclase is the primary phase that incorporates Sr, any melt that forms in
the absence of plagioclase will be enriched in Sr with respect to the residue. The presence of
residual garnet, and to a lesser extent, amphibole, will decrease Y concentrations in the melt,
because HREE and Y are highly compatible in garnet.
High Sr/Y signature (Fig. 2.21) is neither unique to slab melting nor insensitive to the residual
minerals assemblage and fractionating phases.
2.5.2 Lithopheric Cross Sections and the Tectonic and Magmatic History in southern
Peru
Questions of slab depth and crustal sources of 14-16°S (Section 1) and 16-18°S (section 2)
regions are addressed below in relationship to the series of lithospheric sections in Fig. 2.29,
2.30 and 2.31. The frameworks of these sections are based on: 1) present-day CVZ analogues
according to the premise that similar magmas and structures evolve in similar setting; 2)
Geological constraint from an arc to backarc transect through southern Peru (Fig. 2.30).
Crustal thicknesses, lithosperic asthenospheric boundaries and Moho in the CVZ analogues
are based on geophysical models in Yuan et al. (2002), Tassara (2005), and Beck and Zandt,
(2002) and subducting-plate geometries are from Cahill and Isack (1992), Gutscher (2002)
and Hampel (2002). Time slices are chosen to depict sections during and after crustal
thickening at ~29-15 Ma, 10-3 Ma and <3 Ma.
Fig. 2.29. Three dimentional density model of the Andean continental margin from Tassara et al. (2006). a)
Lithosphere-asthenosphere boundary countoured every 20 km. b) Moho every 10 km. Dotted lines depict the
intersection of the corresponding discontinuity with the subducted slab. White lines are section 1 and 2.
53
Fig. 2.30. Eocene, Oligocene and Neogene volcanic arcs in the Western Cordillera and northwestern Altiplano of
south Peru. 1 and 2 are section lines.
Section for the Upper Oligocene to Lower Miocene (Tacaza arc ~29-15 Ma, Fig. 3.31 A)
Section 1, A (14-16°S) Section 2, A (16-18°S)
• Subducting Nazca plate relatively steep. • Subducting Nazca plate steep.
• High Pb isotope compositions. • Low Pb isotope compositions.
• The Uplift of the Western Cordillera began at about 28 to 15 Ma (Sempere et al., 2004;
Wörner et al., 2000).
• The foreland basin system is progressively driven by the uplift and the propagation of a
newly formed Cordillera-the Eastern Cordillera (EC), which started at ~25 Ma (Lamb et al.,
1997). This EC resulted from the west-dipping continental subduction of the Brazilian shield
under the Altiplano.
• Increased plutonic activity at ~20 Ma in the Altiplano (Colque massive, Mamani et al.,
2004) fit with facilitation of magma storage in the crust in this contractional regimen.
• The Tacaza arc has broadened eastward, volcanism has spread across the Western
Cordillera and northwestern Altiplano as shown by the Huaylillas and Ocuviri ignimbrites,
Cotahuasi-Tuti and Ayaviri lavas, Pinaya and Tacaza centers. The greater part of the
metallic mineralization (Cu-Ag-Pb) is hosted by Tacaza volcanics rocks.
• Increased ignimbrite activity at ~24 Ma and ~14 Ma in the frontal arc (e.g. Huayllillas,
Tosdal et al., 1981) and in the backarc (e.g. Ocuviri, Mamani et al., 2004) indicate uplift of
these areas. Thickening of the Asthenospheric wedge over the steepening subducted slab
would have brought in hot asthenospheric melts, facilitating extensive melting of the
previously hydrated continental lithosphere. Injection of these mantle-produced magmas
into the hot, thickened crust would cause massive crustal melting, producing the magmas
that subsequently erupted as large ignimbrite sheets.
• Relative flat REE pattern in the Tacaza volcanic mafic magmas being due to partial
melting of garnet-free mantle. Subsequent fractionation under low-pressure conditions is in
accord with fractionationg assemblages dominated by olivine, plagioclase, and pyroxene
(Mamani et al., 2004).
• Low 87Sr/86Sr ratios high εNd values fit with minimal crustal contamination in a relative
thin crust.
54
Section for the Middle to Upper Miocene (Lower Barroso arc ~10 - 6 Ma, Fig. 3.31 B)
Section 1, B (14-16°S) Section 2, B (16-18°S)
• Crust ~70 km thick. • Crust ~70 km thick
• Slab dip as somewhat shallower than in • Subducting Nazca plate steep
the Oligocene. Shallowing fits with waning • During 7 and 6 Ma andesitic to dacitic
of magmatism at 14°S as the arc broadened volcanism migrates into the Western
southeastward from ~10 Ma to 3 Ma. Cordillera, period of widespread volcanism.
• The emplacement of the Shila, • Small centers are preserved as small
Orcopamapa silver and gold deposits (Fig. stocks and domes toward west of Titicaca
3.32) is linked to magmatism processes at Lake.
~10-9 Ma (Tosdal et al., 1999; Cassard et al., • Low Pb isotope compositions like the
2000). Tacaza arc.
• High Pb isotope compositions like the
Tacaza arc.
• Voluminous ignimbrite sheets, with ages ranging from 10 to 5 Ma, were erupted around
Caravely, Pausa, Aguada Blanca (Wörner unpublished data) indicating renovation of crustal
thickening beneath the arc during Upper Miocene.
• Intense shortening in the sub-Andean zone initiated in the mid- to late-Miocene (10–20Ma)
with >100 km of underthrusting of the Brazilian shield (Lamb et al., 1997). Deformation style
is characterized by the development of a series of thrust imbricates involving Paleozoic
sedimentary rocks (Gil, 2001).
• Crustal thickening is also consistent with the high Sm/Yb ratios that reflect retention of
middle and, to lesser extent, HREE could be in residual amphibol or garnet.
• Relatively minor increase in 87Sr/86Sr ratios and decreasing in εNd values in ca. 10-7 Ma
(Fig. 3.23) such relatively small isotopic range require little change in the type or amount of
crustal contaminat as the sub-Andean crust thickened in this period.
Section for the Pliocene to Recent (Upper Barroso and Frontal arcs < 3 Ma, Fig. 3.31C)
Section 1, C (14-16°S) Section 2, C (16-18°S)
• In the last ~3 Myr subducted slab is shown • The subducting slab is again steep in
shallower than at ~11 Ma, due to the accord with the modern Benioff zone
subduction of the Nazca ridge (Hampel, geometry under the CVZ (Cahill and Isack,
2002). Shallowing fits with waning of the 1992).
magmatism. • Frontal-arc migration to the westward
• High Sr concentration and steep REE from 3 to 0.08 Ma. The actual arc stabilized
pattern (Fig. 3.21) in younger lavas 30 km to the west by ca. 0.08 Ma, and is at
“adakites” used as evidence for slab melting 250 km distance from the trench.
by Gutscher et al. (2000). • Low Pb isotope compositions like the
• High Pb isotope compositions like the Tacaza and Lower Barroso arcs.
Tacaza and Lower Barroso arcs.
• Crust >70 km thick (Yuan et al., 2000). The present-day configuration of the Central
Andes was acquired at ~11 Ma (Lamb et al., 1997).
• Major Pliocene to Recent volcanic activity is essentially restricted to the Western
Cordillera. Using the geometry of Cahill and Isacks (1992) model, the Frontal arc are about
120 to 150 km above the slab and is underlain by a lithosphere thinner than 80 km. Backarc
volcanic activity is limited to the shoshonites along the Vilcanota-Ayaviri lineament fault.
• Sencca ignimbrites with ages from 2.7 to 3.2 (Wörner et al., 2000 and unpublished data)
erupted principally in the Western Cordillera.
• The particular high La/Yb and Sm/Yb ratios, and high Sr contents in the young volcanics
55
fit at least in part, with processes in wich garnet residual of thick crust (Kay et al., 1999).
Aditional support for interaction of magmas with thickening garnet crust at high pressure
comes from the flat REE pattern.
• Increases in 87Sr/86Sr ratios and decreases in εNd values in Upper Barroso and Holocene
magmas stand in merked contrast to large differences between Lower Barrosos and Tacaza
magmas. Such isotopic range requires change in the amount of crustal contaminant.
Geological evidence for crustal thickening correlate also with trace elements, high Sm/Yb and
Sr/Y.
• Geophysical studies suggest that virtually all of the mantle lithosphere below the eastern
Altiplano and western part of the Eastern Cordillera has been removed (Myers et al., 1998;
Dorbath and Granet, 1996). The relatively strong Brazilian lithosphere is underthrusting as far
west as the high elevations of the western part of the Eastern Cordillera (65.5°W) but does not
underthrust the entire Altiplano. As the Brazilian craton underthrusts the subcrustal
lithosphere (and possibly the mafic lower crust), the subcrustal lithosphere and the lower crust
are decoupling from the upper crust and delaminating or subducting into the mantle. The
subcrustal lithosphere is delaminating piecemeal and is not completely removed beneath the
central Altiplano (Fig. 2.32: Beck and Zandt, 2002).
Fig. 2.31. Sequence of schematic lithospheric-scale cross sections showing the Tacaza to Recent magmatic arcs
and tectonic evolution near 14-16°S (Cross Section 1) and 16-18°S (Cross Section 2). The framework of each
section is based on a modern analogue along the northern CVZ. Frontal volcanic centers (triangles) active at the
time shown in the section are plotted at their present distance relative to the modern trench (km 0 on the
horizontal scale). Active volcanic centers, plutons, and faults are schematically shown for each time. The
distributions of faults, foreland-basin sequences are based on GIS Andes (from BRGM). Fragment of crust and
lithospheric mantle from the forearc are shown entering the mantle wedge under the arc at times of arc
migration.
56
Fig.2.32. Schematic cross section showing interpretation of the lithospheric structure of the central Andes from
geophysical and geological studies (Beck and Zandt, 2002). Red and blue indicate upper mantle P wave
velocities that are slower and faster, respectively, than the reference IASPEI-91 model (Myers et al., 1998).
2.6 Conclusions
Oligocene to Recent volcanic centers in the northern CVZ (14-18°S) reveal systematic
differences in trace elements and Sr-Nd isotope ratios with age, which reflect different
degrees of crustal contamination. High-K, calc-alkaline andesites are the most common rock
type in the 36 volcanic centers sampled. The major elements and most trace elements show no
differences with age when andesites of the same degree of fractionation are compared.
However Sr shows striking enrichments in the Upper Barroso to Recent andesites (<3 Ma)
relative to their older equivalent (Lower Barroso and Tazaca rocks).
Regional differences exist: Barroso and Recent volcanic rock erupted north of 16°S have
higher εNd values, low Sr isotope ratios and higher Pb isotope ratios than volcanic erupted to
the south of 16°S (with lower εNd values, higher Sr isotope ratios and lower Pb isotope
ratios). These support the notion that the magmas are controlled by the composition of the
Andean crust in southern Peru.
All younger volcanoes in the northern CVZ have high Sm/Yb and Sr/Y ratios consistent with
residual garnet in the source region. This excludes an “adakitic” signature and any
involvement of slab melts of the volcanic rocks located between 16°S and 14°S.
The composition of andesites in the northern CVZ record a major change between 10 to 3 Ma
after establishment of the thick crust and uplift of the Andes.
57
3 Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the Central

Andes (13°S to 28°S)
Abstract
Meso-Cenozoic magmatic arc systems of the Central Andes (CA) result from the subduction
of the Nazca Plate beneath the South America plate since Jurassic time, the arc progressively
shifted ~180 km from a western position in the Jurassic to the present Western Cordillera in
the Oligocene time. These magmas formed prior, during and after crustal thickening. Present-
day continental crust reaches a thickness of >70 km. Enhanced uplift between c. 30 and 15
Ma is documented by large clastic wedges in Western Andean Escarpment (Wörner et al.,
2000; Sempere et al, 2004). The upper crust is formed by Precambrian to Paleozoic
metamorphic rocks and sedimentary marine Mesozoic rocks, covered by mostly continental
sediments of the Cretaceous and Tertiary.
Evolutionary pattern in geochemistry, initial Sr- Nd isotope and present day Pb-isotope
characteristics of Central Andean (200 Ma to Present) magmatic rocks suggest that the pattern
in Meso-Cenozoic magmatic evolution in this region was controlled by crustal thickness and
the local basements.
3.1 Introduction
Geochemical and isotopic variations of modern Andean magmatism arc and rocks tend to
correlate with changing crustal thickness through time in the Central Andes (e.g. Kay et al.,
1994, 1999; Davidson et al., 1991; Trumbull et al., 1999) and along the arc in the Southern
Volcanic Zone (Hildreth and Moorbath, 1988). These correlations assume that melts
equilibrated with and reflect the source mineralogy of a hydrous mafic melting assimilation
storage homogenization (MASH) (Hildreth and Moorbath, 1988) crust-mantle boundary and
not with variably enriched asthenospheric mantle due to contamination by subducted
sediments or tectonic erosions suggested by Stern (1988, 1990) and/or partial melting of the
enriched sub-continental lithospheric mantle triggered by slab derived fluids proposed by
Rogers & Hawkesworth (1989). Crustal thickening result in increasingly higher pressures for
the residual mineralogy of melts produced from this source region. Geochemically, most
diagnostic of pressure-induced changes in the source mineralogy, from lower-pressure
assemblages dominated by plagioclase and clinopyroxene to higher-pressure amphibole and
garnet-bearing mineral assemblage passing through the garnet stability field at depths >12
kbar or crustal thickening >40 km (e.g. Kay et al., 1991, 1994, 1999; Petford et al., 1996;
Haschke et al., 2003).
Since the beginning of subduction-related Andean magmatism some 200 m.y. ago, repeated
stepwise eastward shifting of the main axis of arc magmatism generated a collage of four
roughly parallel, eastward younging Andean magmatic arcs (Coira et al., 1982; Scheuber et
al., 1994): 1) The Early Jurassic-Early Cretaceous arc along the present Coastal Cordillera; 2)
The Middle Cretaceous arc aligning with Longitudinal Valley; 3) The Cretaceous-Eocene arc
largely corresponding to the western edge of the Western Cordillera; and 4) The Late
Oligocene-Late Miocene arc and modern CVZ composing the Western Cordillera and
Altiplano-Puna plateau, where occur the greatest chemical variations. Each set of arcs have or
no their magmatic backarc to ~120 km distance. These spatial and temporal variations make
the Central Andes an ideal zone to studying long term (~200 Ma) crustal evolution and
magmatic processes at subduction-related.
I compiled major and trace elements and isotope data for these successive arcs between 13°S
and 28°S and found similar and different evolutionary pattern in geochemistry with time and
space.
58
Each suite of arc magmatic rocks is characterized by increasing of Sr-Nd isotopic through
time, Sm/Yb ratios increase with time, but with variable ratios along CVZ. Pb-Nd isotopes of
rocks are not similar along the Central Andes, but are similar at any given area (e.g. in the
central CVZ, Wörner et al., 1992; Aitcheson et al., 1995 and in the southern CVZ, Lucassen
et al., 2001, and in northern CVZ, Chapter 2 of this thesis). Such patterns along the Central
Andes were neglected and previous studies and discussions were concentrating in local areas
(Fig. 3.1). To understand and use these patterns along the entire Central Andes require also
study regionally in light of what control geochemical variations in time and space.
3.2 Database
From the literature I compiled major and trace elements and Sr-Nd-Pb isotopic data (583
samples) for Central Andean magmatic arc rocks (200 Ma to present). Data for Jurassic to
Early Tertiary rocks (Boily et al., 1989; Mukasa et al., 1986; Anthes, 1993; Guerra, 1993;
Rogers and Hawkesworth, 1989; Haschke et al., 2002; Lucassen et al., 2002; Kay et al., 1994;
Horn, 1992; Heumann, 1993). For late Oligocene to present magmatic rocks data come from
Davidson et al. (1990), Horn (1992), Kay et al. (1994, 1999), Rogers and Hawkesworth
(1989), Eusterhues (1993), Aitcheson et al. (1995), Wörner et al. (1988, 1994 and unpublished
data), Mamani et al. (2004), Kohlbach (1999), Trumbull et al. (1999), Siebel et al. (2001). I
also present 201 new analyses for major and trace elements as well as a large subset for Sr-
Nd-Pb isotopes of volcanic and intrusive rocks with ages between 190 and 0 Ma, to fill spatial
and temporal gaps and complete isotope analysis from earlier work and to establish the
pattern along the Central Andes (Fig. 3.1). Sample locations, compiled data and representative
new analyses are listed in database. The complete database for CA and map location of
samples is available on CD.
59
Fig. 3.1. Simplified geological map of the Central Andes. Different square colours indicate the zones of previous
studies.
60
3.3 Geochemistry of Meso-Cenozoic magmatic rocks in the CA
3.3.1 Major and trace elements

Volcanic and intrusive rocks from the Central Andes (CA) can be subdivided into members of
two major magma series: alkalic and sub-alkalic. This distinction is presented in the Total
Alkalis versus Silica (TAS) diagram. The alkalic serie is represented by backarc rocks, and
the samples from Neogene arc and Jurassic-Early Tertiary intrusions plot in the transition
zone between sub-alkaline and alkaline (Fig. 3.2).
In a TAS-diagram, compositional classification of the magmatic samples was done according
to the international nomenclature (Wilson, 1989). Intrusive rocks of the area are also plotted
into the diagram. Calc-alkaline samples are predominantly the Neogene lavas, varying in
composition between basalt andesite, andesite, dacite and rhyolite. Backarc volcanic samples
are “basanites”, a less alkalic suite ranges from basalt to basaltic-trachyandesite. Samples of
the Jurassic and Early Tertiary rocks from the forearc are gabbros or basalts, andesites or
diorite, granodiorites and granites.
Fig. 3. 2. TAS diagrams from Wilson (1989) of own and compiled data from the CA.
Harker diagrams show the chemical differentiation or mixing trend of the CA rock series in
respect of the major elements and some plots with trace elements are added. Besides trends
due to fractional crystallisation of different mineral phases, mixing lines are also observed.
Samples of the Jurassic intrusions are the most primitive of the CA, i.e. they have undergone
the least crystal fractionation. This is obvious from the element variation diagrams. The
Oligocene alkaline rocks from the backarc are affected by some olivine and clinopyroxene
fractionation. Few Jurassic intrusions show deep fractionations of these two minerals.
The most mafic rocks are Oligocene and Recent rocks from the backarc and some Jurassic
rocks from the Coastal Cordillera with characteristically high MgO (> 4 wt %), Ni (> 30
ppm), and Cr (> 100 ppm) contents, but still lower MgO contents compared to primitive
mantel basalt (Fig. 3.3).
Plots of CaO, SiO2, and Al2O3 versus MgO present a marked inflection in the liquid line of
descent that is related plagioclase fractionation or the generalities of a mixing process. At high
61
MgO values only Ol and Cpx fractionation occurs (e.g. backarc samples), whereas around 4%
MgO, plagioclase enters the fractionating assemblage.
Samples of the Neogene lavas and Cretaceous intrusions have multiple mixing.
Fig. 3.3. Element variation diagrams for major and trace elements versus MgO. Folds in the liquid lines of
descent report the start of crystal fractionation of a new mineral facies.
Harker diagrams of major element oxides versus SiO2 and some additional diagrams of trace
elements versus SiO2 (Fig. 3.4) characterise the major fractionating phases olivine,
clinopyroxene, plagioclase, as well as alkali-feldspar, magnetite/sphene and the accessory
phases like apatite and zircon and mixing process.
62
Fig. 3.4. Harker diagrams showing elemental variations of major elements and some trace elements versus SiO2.
The rock groups from the CVZ display liquid lines of descent with clear evidence for fractionation other straight
lines due to mixing. ol=olivine, cpx=clinopyroxene, plag=plagioclase, mt=magnetite, il=ilmenite, fsp=feldspar.
In generall ratios of major elements of rocks with >57% SiO2 are governed by multiple
mixing between shallow magmas (<57% SiO2) and more evolved end of “deep”magmas as it
showed in Rb versus Ni diagram (Fig. 3.5a).
In the SiO2 (wt%) versus U and Th diagrams is observed the ignimbrite fields (Fig. 3.5b, c),
which indicate shallow level fractionation of accessories in this rocks. The trends follow the
mixing between the deep fractionation magmas and more evolved magmas.
63
Fig. 3.5. a) Diagram Rb (ppm) versus Ni (ppm), b) and c) SiO2 (wt%) versus U (ppm) and Th (ppm) to
distinguish between deep and shallow fractionation and mixing process. Legend as in Fig. 3.4.
As we observed in the Figures 3.3, 3.4 and 3.5, geochemical data shows that arc lavas with
SiO2 (55 to 65 wt %, Fig. 3.5) content are product of mixing, this could explain why in the arc
lavas there are no mafic lavas with ~50 % SiO2 and most of the lavas have > 55% SiO2 (Fig.
3.6).
Fig. 3.6. Histogram of SiO2 (wt %) concentration and Sr (ppm) content for samples from the CA.
64
Spider diagrams
Mantle-normalised (McDonough et al., 1992) multi-element diagrams are illustrated for zones
along the CVZ in (Fig. 3.7). The patterns show typical subduction signatures for almost all the
samples. HFSE (Nb, Ta, Hf, Zr, Ti) depletion is a characteristic for this tectonic environment
and most conspicuous with the Nb-Ta trough. This trough is most clearly visible in samples
from the Jurassic-Tertiary intrusions and Neogene lavas.
The patterns of the Jurassic and Cretaceous are definitely distinct in that they show a more flat
signature with almost flat REE.
Completely contrary is the pattern of the Neogene lavas. The pattern of the Mio-Pliocene to
Holocene have a steep gradient from La to Y, in comparison Miocene are intermediate.
Sr concentrations in Neogene lavas are different e.i. high in zone 1 and 2, and low in zone 3.
The patterns of the backarc samples (shoshonites) show less depletion of the HFSE,
enrichment of fluid-mobile elements (Ba, Th, U) and La, Ce, Sm, Tb in comparison with
other rocks. Ba and La content are lower in backarc arc lavas, Ta are depleted as the Holocene
lavas and not enriched as the backarc lavas. This is demonstrated in the diagram by
comparison to the pattern of OIB (Fig. 3.7).
Different to the lavas, the Neogene ignimbrites do not display strong Ta-Nb depletions, but
have strong negative Sr anomalies and Ba troughs due to intensive feldspar fractionation. The
ignimbrites from the Zone 1 exhibit positive Ba anomalies and less negative Sr anomalies
suggesting that no K-feldspar fractionation occurred. Depletion at Ti results from
fractionation of accessory phases such as magnetite and sphene.
Rare earth elements (REE)

A good tool for the distinctions of magmatic rocks from Central Andes is REE plots (Fig.
3.8). The REE contents of the samples are normalised to chondrite concentrations after
McDonough & Sun (1995). For comparison, pattern from Jurassic to recent arc of the CVZ
are shown in the diagrams.
The Jurassic samples display a pattern with little enrichment of the LREE (La/Sm), while the
Cretaceous rocks pattern is to some respect comparable but shows in most cases a negative Eu
anomaly due to plagioclase fractionation, and higher LREE concentrations.
The REE pattern of the Neogene lavas is definitely distinct: the Miocene samples show a
signature with moderate gradients from LREE to HREE. Completely contrary is the REE
pattern of the Mio-Pliocene to Holocene samples, because their patterns are steeper. The
gradient variations of the Neogene rocks change from zone 1 (steeper) to zone 3 (lest steep).
LREE pattern for the ignimbrites from the three zones are similar and characterized by a large
range. Eu and HREE patterns are different, the marked negative Eu anomalies and HREE
contents are lower in the ignimbrites from the Zone 2 (Fig. 3.8).
The Shoshonites LREE patterns are different from the other samples of CVZ. These young
backarc volcanics display a generally steeper LREE pattern than arc and forearc and the
HREE pattern are relatively flat.
65
Fig. 3.7. Primordial mantle normalized concentrations of trace elements plus K and Ti from northern, central and
southern part of the Central Volcanic Zone. Normalization factors are from McDounough et al. (1992). The OIB
pattern (Sun, 1980) is shown for comparison.
66
Fig. 3.8. REE diagrams for the CVZ. Normalization factors are from McDonough & Sun (1995).
3.3.2 Isotopes
Sr- and Nd-isotopes
The initial isotopic composition of the rocks provides information about the (mantle-) sources
of the magmatic rocks as well as about processes by which their chemical and isotopic
compositions are modified (e.g. crustal contamination).
When the age of the rock is known, its initial isotopic ratios can be calculated by using the
basic geochronological equations:
67
87
Sr ⎛ 87 Sr ⎞ 87
Rb λt
=⎜ ⎟ + (e − 1)
86
Sr ⎜⎝ 86 Sr ⎟⎠ initial 86
Sr
which are transformed to:
⎛ 87 Sr ⎞ 87
Sr 87
Rb λt
⎜⎜ 86 ⎟⎟ = 86
− 86
(e − 1)
⎝ Sr ⎠ initial Sr Sr
The epsilon notations (DePaolo & Wasserburg 1976) are calculated by using the following
equations:
⎡ (143Nd / 144Nd ) initial ⎤
ε Nd = ⎢ t
− 1⎥ * 10 4
⎣ I CHUR ⎦
Jurassic and Early Tertiary intrusions (data from Boyle et al., 1989; Mukasa et al., 1986;
Hawkesworth et al., 1989; Clark et al., 1990; Anthes, 1993; Lucassen et al., 2002) show
distinctly positive εNd and lower 87Sr/86Sr than younger samples, most of the intrusions
overlap with the ratios reported from NVZ, SVZ and AVZ.
Fig.3.9. Diagram of εNd vs. Initial 87Sr/86Sr presenting the data set from the CVZ in comparison to data fields
from NVZ (Bourdon et al., 2002), SVZ (Kay et al., 2005), AVZ (Stern and Killian, 1996). BSE is bulk silicate
earth.
The isotopic variability of the Neogene magmatic rocks have values ranging between
87
Sr/86Sr= 0.705 to 0.7085 and εNd= 0 to -12. Jurassic, Cretaceous, Eocene rocks are more
depleted (e.i., higher εNd and lower 87Sr/86Sr) than those of the Neogene arcs, and overlap the
ratios reported from SVZ and NVZ.
Shoshonite samples display less negative εNd and middle 87Sr/86Sr ratios compared with
cluster formed by all volcanic rocks of the CVZ.
Oligocene lavas are less depleted in εNd = 0 to -7 than Neogene lavas.
The ignimbrites show a wide range of εNd and higher 87Sr/86Sr ratios, some of them overlap
the Neogene lavas.
68
Spatial and Temporal variation of Sr-Nd isotopes

As see in Fig. 3.9 87Sr/86Sr ratios increase with decreasing age from Jurassic to Miocene.
During Mio-Pliocene until present, Sr-isotopic compositions remain nearly constant with most
ratios falling within a range of 0.705-0.7085. Mio-Pliocene to Holocene volcanoes displays
more variable values than Miocene, probably simple reflecting the larger number of available
data.
Fig. 3.10. 87Sr/86Sr and εNd ratios versus Latitude °S.
Increasing 87Sr/86Sr from the coast to Western Cordillera with decreasing age has been
observed by Rogers & Hawkesworth (1989) at 22°S, Kay et al. (1994) at 26-29°S and
Haschke et al. (2002) at 21-26°S. In all work cited, Mio-Pliocene to recent samples show
drastically elevated and variable Sr-isotope ratios, probably reflecting increased crustal
thickness and assimilation resulting from the recent uplift episode.
Nd-isotopes ratios decrease with decreasing ages. Jurassic and Early Tertiary rocks drop from
εNd 5 to εNd -2.
The Eocene, Oligocene and Neogene rocks delineate three groups for Nd-isotope ratios. In the
central sector between 15°S and ~25°S, εNd ranges from -4 to -9. The notable low εNd values
from -9 to -12 are around 16°S (Arequipa). The northern sector (13°S - 16°S) and southern
sector (24°S-28°S) have similar values of εNd from 0 to -6. Transitions between these groups
are relatively abrupt.
The Mio-Pliocene to Holocene volcanoes between 16 to 22 °S show notably high 87Sr/86Sr
ratios and low εNd values (Fig. 3.9) than the volcanoes from (13- 16°S and 22- 28°S), and high
69
Sr contents (400 to 1600) are observed between 22 and 14°S (Fig. 3.11), although crustal
thickness is similar along these latitudes.
Concentration between 0 and 400 Sr (ppm) is conformed only by ignimbrites (SiO2> 60%),
Jurassic and Cretaceous rocks.
Fig. 3.11. Sr contents along the central volcanic zone. Legend as in Fig. 3.10.
Spatial and Temporal variation of Pb- isotopes

New lead isotopic compositions (68 samples) were determined on carefully selected whole
rock and were combined with published data.
The majority of the analyses are Neogene lavas. En general the Meso-Cenozoic samples show
two groups that are distinct in their 206Pb/204Pb ratios (Fig. 3.12). The low 206Pb/204Pb =17.5-
18.5 and the high 206Pb/204Pb =18.5-19. The rocks from the basements also show two groups:
Paleozoic rocks have the highest 206Pb/204Pb > 18.5 and plot in the radiogenic site, and
Proterozoic samples show the lowest 206Pb/204Pb < 18 (unradiogenic site).
In the diagram 207Pb/204Pb versus 206Pb/204Pb, the data set plot between to the upper crustal
lead and orogenic belt lead of Zartman and Haines (1988).
Jurassic intrusion plot near the Nazca plate basalts field and Nazca plate sediments in the
diagram 208Pb/204Pb versus 206Pb/204Pb.
The different Pb-isotope ratios observed in the Cenozoic rocks derivate from the isotopic
compositions of the underlying basements, the low Pb ratios overlap the Proterozoic points,
while the high lead isotope overlap the Paleozoic samples (Fig. 3.12). Such tendency of the
206
Pb/204Pb found in volcanic rocks has been interpreted as a result of mixing between mantle-
derived Pb and Pb crustal by various authors (e.g. Tilton & Barreiro, 1980; Barreiro, 1984;
Mukasa et al., 1986; Wörner et al., 1992; Aitcheson et al., 1995).
Based on assimilation model for some volcanoes of the CVZ, crustal input about 12 and 20%
into mafic mantle magmas through melting in the middle or lower crust prior to
differentiation in the upper crust (Aitcheson et al., 1993 and Entenmann, 1994).
Some Jurassic and Early Tertiary intrusion are interpreted to reflect either variations from a
more radiogenic mantle source or assimilation of rarely more radiogenic or unradiogenic
crustal components.
70
Fig. 3.12. Pb-isotope diagrams of the Meso-Cenozoic rocks from the CVZ, compared with the lead isotopes
composition from the Proterozoic and Paleozoic basements (outcrops in Fig. 1.3), Nazca Plate sediments and
basalts (Dash, 1981). Zartman and Haines (1988) lead evolution model: M. upper mantle lead; O. orogenic belt
lead; U. upper crustal lead; L. lower crust lead. The ticks on each curve are at 100 Ma intervals.
Although the overlapping of all samples along the CVZ (Fig. 3.13), three groups independent
of age and SiO2 contents, show abrupt transition at 18.5 to 18.7 ratios in the diagram
206
Pb/204Pb versus Latitude °S. In the central part between 15.30°S and 20.8°S 206Pb/204Pb
varies from 17 to 18.50. The high 206Pb/204Pb ratio of 18.5 to 19.5 is found in 13-16°S and 20-
28°S. Such variations are correlated with Nd-isotopes variations along the CVZ (see Fig.
3.10) and with Sr-isotopes are not clear the correlation.
In the diagrams 207Pb/204Pb and 208Pb/204Pb, the ratios show no steep transitions as the
206
Pb/204Pb along the arc.
71
Fig. 3.13. Regional variation in Pb isotope compositions along the CA.
3.4 Summary and Discussion: Implications for a tectonic model
Summary
Following observations have been made along the CA.
• Mesozoic intrusions and Cenozoic lavas from the arc are classified calc-alkaline and
Cenozoic rocks from the backarc are alkaline. These rocks display signatures characterized by
LILE enrichment and HFSE depletion. The ignimbrites do not show a marked Nb-Ta
depletion.
• Concentrations of, P2O5, K2O, Na2O, Sr, Ba, Zr, Nb increase from Jurassic to present,
whereas abundances of, CaO, Fetot and MnO decrease. Content of Y are highest in
Cretaceous-Eocene rocks and remain at constantly low values from Mio-Pliocene to Present.
• The slope of REE patterns increase from Jurassic to Holocene rocks, as Yb contents
decrease and Sm contents increase and become more variable with time.
72
• (87Sr/86Sr)initial ratios increase from Jurassic until present.

• 143Nd/144Nd ratios decrese from Jurassic until present.
• Pb isotope compositions of igneous rocks clearly reflect a contribution of the underlying
basement.
Discussion
Geochemical variations across or along the active continental Andean margin have been
investigated by numerous authors (Fig. 3.1, Kay et al., 1994, 1999, 2005; Stern, 1990, 1991;
Hildreth & Moorbath, 1988; Roger & Hawkesworth, 1989; Wörner et al., 1988, 1992;
Davidson et al., 1990; Haschke et al., 2002, 2003). The uncertainties about the nature of this
complex subduction setting characterized by an interplay of many components (such as
descending slab, sediments, fluids, asthenospheric mantle, lithospheric mantle lower and
upper crust) by various processes (partial melting, assimilation, fractional crystallization) led
to the development of conflicting models on Andean Petrogenesis.
The following discussion try to assess debate different sources or different processes are
responsible for the observed variations in major, trace elements and isotope geochemistry
along the CVZ.
3.4.1 Source or Processes?

As described in Chapter 1 and in the Introduction, the large number of conflicting models can
be split into two major groups: One group proposes that the geochemical characteristics of
erupted lavas simply reflect heterogeneity of the source. A heterogeneous source either
represents variably enriched asthenospheric mantle due to contamination by subducted
sediments or tectonic erosions (Stern, 1988, 1990) or ancient enriched lithospheric mantle
(Roger & Hawkesworth, 1989).
The second group postulates that all primary magmas produced at subduction zones are
essentially similar (since they have been derived by similar mechanisms from a similar
source, the asthenospheric mantle wedge) and inherit their diverse geochemical features by
interaction with the crust during ascent (Hildreth & Moorbath, 1988; Wörner et al., 1988,
1992; Davidson et al., 1990).
Subducted sediments & Tectonic erosion

Although Stern (1988) holds variable subduction geometry and sedimentary input responsible
for along arc variations in magma geochemistry between 33°S - 38°S, Hildreth & Moorbath
(1988) have demonstrated that rather varying amounts of crustal contributions can account for
the observed diversity. The question is, despite of apparent constancy of sedimentary input
along arc today, whether changes in subduction geometry and/or sedimentary input through
time have influenced magma geochemistry across arc. Seyfried et al. (1995) recognized
crucial changes in the drainage system and humidity regime at 18°S from Miocene until
present: in contrast to the present extremely arid climate impeding intense erosion, over 1000
m of alluvial sediments were deposited in the Longitudinal valley (Fig. 1.2) during the
Miocene times. Flores et al. (2005) suggested increasing subsidence and synsedimentary
activity in the Longitudinal valley of southern Peru during the Early and Middle Miocene.
Allmendinger et al. (2005) concluded that it is doubtful that significant amounts of
terrigenous sediments reach the Peruvian-Chilean trench as the Costal Cordillera scarps
prevents high sedimentary input.
A strong influence of pelagic sediments is also not observed in the rocks of this study as no
sample displays negative Ce-anomalies or Pb-compositions resulting from hypothetical
mixing of primitive island arc basalt with Nazca plate sediments (Davidson et al., 1990).
Entenmann (1994) however, demonstrated that mixing of MORB with ~1% pelagic Pacific
73
sediments can account for the variable LILE content of CVZ basaltic andesites. On the other
hand he argued that elevated δ18O values are caused by intra-crustal assimilation and are not
source effects resulting from sediment subduction and tectonic erosion, or derivation from
ancient subcontinental mantle lithosphere.
Plots of Rb/Cs and Ba/La versus age (Fig. 3.14) underlie the apparently conflicting evidences
for or against source contamination: whereas maximum Rb/Cs ratios (low ratios indicate
involvement of Cs-rich pelagic sediments, Plank & Langmuir, 1988) decrease from Miocene
to present, Maximum Ba/La ratios (high ratios indicate involvement of subducted sedimentary
components, Kay, 1977) also decrease or remain constant, contradictory to what one would
expect, if these features were stablished only by increasing amounts of a slab component.
Rb/Cs ratios have their maximum and largest scatter in Miocene times exactly when the uplift
event occurred. As Hildreth & Moorbath (1988) proposed that intra-crustal contamination is
the dominating process in establishing Ba/La ratios, high Rb/Cs and Ba/La ratios migh reflect
the onset of large scale crustal involvement in the 20 Ma to Recent rocks. However, Ba/La
ratios are very sensitive to K-feldspar fractionation, as we see in the Neogene ignimbrites
ratios (Fig. 3.14). Thus, the Ba/La ratios have to interpret with caution.
Fig. 3.14. Rb/Cs and Ba/La versus age diagrams of samples from the CA.
In the light of the obviously contradictory indications above, it is hard to assess the role of
source contamination in Andean Petrogenesis. With respect to strong evidence from Pb
isotopes data, the influence of source contamination is considered to be small.
74
Enriched subcontinetal lithosphere

Roger & Hawkesworth (1989) postulated the ancient enriched subcontinental lithosphere to
be the main site of magma production, as both incompatible elements abundances (including
Sr concentration) and 87Sr/86Sr ratios increase from west to east at 22°S. In contrast to Roger
& Hawkesworth (1989) εNd values decrease and Sr contents increase with time and are
independent of Ta/Sm ratios, Th/Ta ratios does not unequivocally decrease with higher Ta/Sm
ratios and younger ages (Fig. 3.15), indicating that no large contribution of subcontinental
lithosphere occurred with time.
Nevertheless, Stern et al. (1991), reported highly variable bulk compositions from upper
mantle xenoliths in Patagonian basalts at 52°S thus demonstrating the potential heterogeneity
of subcontinental lithospheric mantle (at least for the SVZ).
Fig. 3.15. Diagrams of Th/Ta, εNd and Sr versus Ta/Sm of samples from the CVZ similar to those from Roger &
Hawkesworth (1989). Legend as in Fig. 3.14.
However, as discussed in Davidson et al. (1991), in the Figure 3.16 Mio-Pliocene to recent
lavas trend show high pressure (where garnet is stable) during depth assimilation, in
agreement with a 70 km thick crust after Andean uplift. Contrasting are the Jurassic,
Cretaceous, Early Tertiary, and Miocene samples, which display much lower Sr contents. The
lower Sr-isotope ratios probably reflect a lesser degree of crustal assimilation involving no
75
garnet, in accordance with a thin crust at these times. The Neogene ignimbrites trend show
higher 87Sr/86Sr ratios and low Sr contents indicating that they added additional crustal
component by contamination of low Sr and high SiO2 magmas late in the differentiation
process with plagioclase stable at low pressure.
Fig. 3.16. Initial 87Sr/86Sr versus Sr of samples from the CA. Symbols and data sources as in Fig. 3.14
Oxygen and Pb isotopes data provide convincing evidence against a dominant role of enriched
mantle lithosphere in Andean rocks at 18°S (Davidson et al., 1991, Fig. 3.1).
Crustal Contamination
Although many uncertainties about composition and structure of the continental crust exist, all
data presented here are most consistent with models involving intra-crustal assimilation in
MASH-zones as proposed by Hildreth & Moorbath (1988), Wörner et al. (1988, 1994),
Davidson et al. (1990, 1991). Geochemical trends and Nd-Sr isotope indicate an increase of
crustal thickness with time and clearly document the Andean uplift event which started ~25
m.y. ago.
Constraint on the amount and timing of crustal thickening and on the source region come
from the temporally changing REE signatures of magmatic arcs formed before, during and
after crustal thickening. In simple terms, the REEs are guides to the pressure-sensitive
residual mineral and melting percentage of the source. These correlations assume that the
melts equilibrated with and reflect the source minerals of a hydrous mafic crust-mantle
boundary region showing melting-assimilation-storage-homogenization (Hildreth and
Moorbath, 1988) processes.
Increasing La/Yb ratios in magma across arcs (with decreasing ages) is widely observed
phenomenon (e.g. Gill, 1981; Torpe, 1979; Kay et al., 1991, 1994, 1999; Hawkesworth, 1988;
Haschke et al., 2002, 2003). Hildreth and Moorbath (1988) also observed an increase of
La/Yb along at 33-37.7°. In all cases, La/Yb ratios are higher on thicker and more mature
crust, thus crustal thickness seems to play an important role in processes that fractionate
HREE from LREE.
Haschke et al., 2002, 2003 observed increasing La/Yb ratios and increasing Sr-Nd isotopic
enrichment with time are characteristic of Neogene magmatic rocks and Pre-Neogene
magmatic episodes between latitudes 21°S and 22°S (Fig. 3.17, Fig. 3.1). These temporal
76
geochemical changes reflect a progressive transfer of melting zone from the upper mantle into
variably thickened garnet-bearing (lower) crust.
Fig. 3.17. Plot of age of Central Andean magmatic rocks vs. La/Yb ratios, showing steepening rare earth element
(REE) patterns within each arc system and with thickening of arc crust (from Haschke et al., 2003).
Our compiled data of different magmatic arcs in the CA show dramatic change in the values
of Sr-Nd isotopes and REE ratios from Jurassic to Recent times (Fig. 3.18).
To quantify the general slope of REE pattern through time along the CA (principally during
the “Andean Cycle”), La/Yb, La/Sm, and Sm/Yb ratios are plotted versus age in Figure 3.18.
As was observed by Haschke et al. (2003) La/Yb ratios increase with time in the CA (Fig.
3.18c). However, LREE ratios (La/Sm) show that they are sensitive to plagioclase
fractionation (see ratios of Neogene ignimbrites in Fig. 3.18d and Fig. 3.19) and HREE ratios
(Sm/Yb) are less affected by plagioclase fractionation (Fig. 3.18e). HREE ratios is signal to
changes in pressure and residual crustal mineralogy and high Sm/Yb ratios is the best
indicator of garnet-bearing in residual mineralogy and because garnet preferentially
fractionates the HREE.
Jurassic rocks (50-70 wt % SiO2) show fractionated but largely overlapping REE pattern (Fig.
3.8). Evolution of Sm/Yb ratios of these rocks shows less increase (Sm/Yb from 1 to 3).
Cretaceous-Early Tertiary before-uplift rocks are characterized by LREE-enriched patterns
(Fig. 3.8 and 3.18d) and Sm/Yb (1-3.5). Eocene lavas show Sm/Yb ratios (from 1 to 3.5).
77
Fig.3.18. Semilog scale plotting the Age (Ma) for CA rocks versus Sr-isotope, eNd values, La/Yb, La/Sm, and
Sm/Yb ratios.
Miocene syn-uplift andesitic and dacitic composition (55-65 wt % SiO2, Fig. 3.19) exhibit
fractionated REE pattern with steeper HREE patterns (Fig. 3.8). Sm/Yb ratios of these rocks
increase through time from mostly low Sm/Yb (1-3) to high Sm/Yb (1.5-5) ratios.
78
Mio-Pliocene after-uplift andesites and dacites (53-68 wt % SiO2, Fig. 3.19) show fractionated
REE pattern (Fig. 3.8) and have higher Sm/Yb (2-7) ratios.
Pleistocene to Recent andesites and dacites (53-68 wt % SiO2, Fig. 3.19) show extremely
fractionated REE pattern (Fig. 3.8) with strongly steepening HREE pattern at this time.
Sm/Yb ratios vary from 2 to 10.
Oligocene-Miocene and Quaternary rocks in the backarc (47-65 wt % SiO2) show also
increasing HREE pattern with decreasing age (Fig. 3.19 and 3.18e).
These systematic changes are due to progressively decreasing Yb (from 3.7 to 0.4 ppm) with
decreasing age (Fig. 3.20).
Fig.3.19. Plot of La/Yb, Sm/Yb versus SiO2 (wt %), rocks with more than 68% are affected by plagioclase
fractionation. Legend as in Fig. 3.18.
Fig. 3.20. Sm/Yb ratios versus Yb (ppm).
Many of these characteristics of the youngest lavas Mio-Pliocene to Recent andesites-dacites

match those of adakites (e.i. primary dacite melts from subducted young oceanic crust (Defant
and Drummond, 1990)). However, geophysical, geologic, isotopic, and age data from our
study suggest that partial melting of the base of thick orogenic crust (~40 km during Mesozoic
time, and ~70 km during Neogene).
79
Sm/Yb ratios not increases similar along the CA even they have the same age and cross the
same thick crust, the highest ratios are between latitudes 14°S and 22°S (Fig. 3.21). However,
there is an obvious decoupling between the tectonic development and its reflection in
geochemistry of erupted rocks: while elevated crustal thickness (~70 km) may have been
established by ~10 Ma, the rapid increase in Sm/Yb ratios occurred some ~3 Ma later.
Observed Sm/Yb ratios of 2-10 associated with 70 km thick crust along the CA correspond
well with Sm/Yb ranging 2-8 reported from Kay et al. (1999) associated with ~60 km crust at
28°S-33°S and Sm/Yb ratios of 3-9 at 34°S -36°S.
Fig. 3.21. Sm/Yb ratios along the Central Volcanic Zone.
Dostal et al. (1977) proposed an alternative model in which steeper REE pattern are produced
by smaller (<5%) degrees of partial melting of an enriched upper mantle source across arc
which also results in the increase of incompatible elements abundances. Lopez-Escobar et al.
(1977) modeled the steep REE pattern of a silicic andesite as product of 3% melting of a
garnet peridotite. However, these processes are not capable of explaining the elevated Sr
isotopes ratios, unless subsequent crustal assimilation has occurred. Hildreth and Moorbath
(1988) show that the stepper REE pattern, elevated abundances of incompatible elements and
increased Sr and Oxygen isotope ratios can be accounted for by assimilation of lower crust.
Furthermore, they do not see evidence why melting of garnet peridotite beneath thick crust
should be a more effective sink for HREE than beneath thin crust.
Sr-Nd isotopic data of arc magmatic rocks show isotopic enrichment with decreasing age
(Fig. 3.18a, b and 3.19), overall there is contrast between Jurassic-Early Tertiary (87Sr/86Sr =
0.703-0.705 and εNd = 6 to -2) and Neogene rocks (87Sr/86Sr = 0.703-0.705 and εNd = 0 to -
12). The major change is seen before and after crustal uplift (Fig. 3.22).
Pb isotope signatures reflect different crust (Wörner et al., 1992; Aitcheson et al., 1995) and
the increasing of Sm/Yb of Pliocene to Recent lavas is observed in the radiogenic and
unradiogenic groups (Fig. 3.22).
80
Fig. 3.22. Plot of 87Sr/86Sr ratios, εNd values (indicator of degree of assimilation) and 206Pb/204Pb ratios (indicator
of different crustal contaminant) versus Sm/Yb ratios (indicator of depth of assimilation). cpx = clinopyroxene,
hbl = hornblende, gar = garnet. Legend as in Fig. 3.21.
3.4.2 Geochemical constraint in a Tectonic model

During Jurassic time, the arc developing at the South American Continent margin immature,
as is documented by the geochemical and isotope characteristic of Jurassic samples. Either the
crust was juvenile characterized by low 87Sr/86Sr ratios and positive εNd values, or the degree
of assimilation of mature crust was very small (e.g. along Coastal Cordillera). Accordingly,
the distinct Pb-isotope composition indicates either derivation from a more radiogenic mantle
source or contamination with basement from the Arequipa basement (Proterozoic) or other
basement (Paleozoic). For the Jurassic arc there is not evidence for crustal shortening or
significant crustal thickening, consistent with overall Sm/Yb (1 to 2.5) ratios.
81
In the Cretaceous there are evidences that indicate some shortening and thickening of the
crust between 90 and 80 Ma (Peruvian unconformity, Scheuber et al., 1994) and the temporal
evolution pattern is constrained in Cretaceous-Early Tertiary rocks with Sm/Yb from 1 to 3.5
ratios. Either increasing degree of crustal assimilation or contamination by more mature
crustal material raised 87Sr/86Sr ratios and lowered εNd values along Longitudinal Valley. The
Pb isotope composition of samples from Linga Yarabamba, Punta Coles, Murmantani suggest
contamination of Arequipa basement (Proterozoic). The geochemical character of samples
from Ocoña, Quebrada Blanca, Cerro Colorado, Cerro Ceusis suggests contamination of
Paleozoic basement.
At about ~25 Ma, uplift rates at the Andean margin accelerate and Neogene arc was
established on thickened continental crust as indicating Sm/Yb (1.5–5) ratios in Miocene
lavas. Ignimbrites flare-up at ~24 Ma characterized by eruption of voluminous ignimbrites
(e.g. Oxaya, Huaylillas) mark episode of major crustal melting. 87Sr/86Sr ratios, εNd values and
Pb-isotope ratios of younger lavas (Mio-Pliocene to Recent) indicate increase in the
assimilation of their basements. Sm/Yb ratios (1.5 to 8) of the 10-3 Ma lavas document crustal
thickening during Mio-Pliocene times, as the MASH zone probably had no thermally
equilibrated yet (see REE pattern).
As the MASH zone approaches thermal equilibration, more fertile rocks are partially molten
in the Mio-Pliocene times.
The younger lavas (<3 Ma) north of 22°S shows dramatically elevated Sm/Yb ratios and
higher incompatible element abundances, which is a stronger indication of interaction with
lower garnetiferous crust. To the South Sm/Yb is lower although crustal thickness is similar.
This difference may then reflect the additional effect of variable crustal composition, mafic in
the north (13°S-22°S) and may be more felsic in the south (22°-28°). Regional differences
exist in other trace element ratios as well.
Perspectives
While the influence of increasingly thick crust and different basement rocks on magma
genesis has been fairly well documented and had compiled an extensive geochemical and
isotopic database of the Central Andes (13°-28°S and 75°- 66°W) are needed a spatial
analysis of the data to better constraint the regional variation of Meso-Cenozoic magmatic
arcs. Geographic Information System (GIS) could help to show the exact locations where the
geochemical variations are. Especially Pb-isotope data would be useful to investigate the
nature and extention of the contaminat basements in the Central Andes.
82
4 Geochemical variations in time and space of lead isotopic domains in the Central
Abstract
Boundaries of the Pb-isotope domains in the Central Andes of southern Peru, northern Chile,
northwestern Argentina and Bolivia are presented based on 356 selected published data and
446 new Pb-isotope analyses of rocks and ores from the Central Andes. A GIS database
contains further geochemical, isotopic, geological, metallogenic, geochronological and
geophysical information that constrain the chemical compositions of rocks, large structural
trends, main mining districts and provinces, as well as, the depth of the Moho and the
subducted slab. Combining Pb isotopes wih Sr, and Nd isotopes further allows to characterize
and distinguish these domains. These data are combined with a 3D density model of the
continental margin of the Central Andes. Crustal domains are thus constrained much more
precisely as previously has been possible boundaries in the Central Andes. Geochemical and
isotopic data in Jurassic to Recent volcanic rocks from these different domains show
increasing crustal assimilation of local basement through time. Combining the isotopic
domains map with 3D density model shows a surprising correlation between the average of
crustal densities and isotopic compositions of the domains. Some domain boundaries are also
enhanced by large crustal scale tectonic structures. Occurrence and type of Cu-Au-Ag-Sn
deposits also appear to be correlated to certain crustal domains.
This evidence suggest that several crustal compositions construes density, crustal structure,
Central Andean plateau segmentation, ore formation and the composition of emplaced
magmas through crustal assimilation.
4.1 Introduction
The Central Andes contain most dense outcrops of Meso-Cenozoic magmatism and
concentrations of ore deposit in South America. Here, the magmatic rocks and ore deposits
were formed on different basement with different geological histories and isotopic
characteristics (James, 1982; Barreiro, 1984; Barreiro and Clark, 1984; Puig, 1988; Gunnesch
et al., 1990; Kontak et al., 1990; Macfarlane et al., 1990; Mukasa et al., 1990; Aitcheson et al.,
1995; Tosdal, 1996; Wörner et al., 1992).
We evaluate >1100 rock samples for mayor and trace elements as well as a large subset for
Pb-Nd and Sr isotopes from the Central Andes, which we combine with selected published
data. The basement in southern Peru, western Bolivia and northern Chile, northwestern
Argentina (13° to 28°S and 75° to 66°W) is heterogeneous: Felsic from 21° to 27°S (Lucassen
et al., 2001). Mafic the northern part of the Arequipa terrane (Tosdal, 1996). The basement
below Altiplano-Puna at 22°S is felsic to the depth of 50-55 km but has more mafic
composition below this depth (Yuan et al., 2002). The only significant juvenile additions to
the crust are confined to the Mesozoic magmatic arcs mainly in the Peruvian and Chilean
coastal range (Mukasa, 1986; Lucassen et al., 1996).
Various studies suggest that lead isotopic compositions of igneous rocks and ores in the
Central Andes reflect the compositions of the underlying basement and thus can be used to
map out crustal domains (MacFarlane et al., 1990; Wörner et al., 1992; Aitcheson et al., 1995)
and for evaluating plate reconstructions (Tosdal 1996; Loewy et al., 2004).
Here, I use a large geochemical and isotopic data (published and new data) to redefine crustal
boundaries at much higher spatial resolution by combining Pb-Nd-, and Sr isotopes and
geochemical trace element data with a 3D density model. This geophysical model (The
geometry of the crustal structure index = Ө, Tassara et al., 2006), which constrain the crustal
domains defined in this thesis.
83
4.2 Information systems and Pb isotope data

I have compiled an extensive geochemical data on the Central Andes (13°-28°S and 75°-
66°W) using a GIS, which contains data and metadata grouped into thus categories:
Geochemistry: a database with >1500 whole rock analysis (major elements, trace elements,
Sr-Nd-Pb isotopes) covering an age range from the Proterozoic to the Present and
encompassing intrusive and volcanic rocks, metamorphic basement and (limited) sedimentary
rocks.
Geology: a geological synthesis at 1:2,000,000 scale and the distribution of Jurassic-
Cretaceous-Paleogene intrusions, Neogene volcanic rocks, and active centers. The compiled
geological map also contains major crustal-scale tectonic lineaments.
Geophysic: gravimetry (Bouguer anomalies), gravity model for the Nazca Plate and the
intracrustal density discontinuity (IDC) geometry and crustal structure index (Ө). Also based
on geophysical data I included depth of Moho and depth contours of the downgoing Nazca
Plate.
Metallogeny: Pb isotope data on 120 Andean ore deposits combined with metadata on
location, main metallogenic features of the mining districts and provinces.
The Principal analytical methods of major elements, trace elements and Sr-Nd-Pb isotopes
carried out on my samples are described in the Appendix.
4.2.1 Lead isotope data

Pb isotope data (Fig. 4.1) are available for Proterozoic basement rocks in only a few scattered
localities in the Central Andes (Fig. 1.3) from the Arequipa basement of southern Peru,
northern Chile and western Bolivia (36 samples from Tilton & Barreiro, 1980; Tosdal, 1996;
Wörner et al., 2000; Loewy et al., 2004 and 16 new samples). Pb isotopes of the Arequipa
basement are unradiogenic with low 206Pb/204Pb = 16.083-18.45, 207Pb/204Pb = 15.606-15.636,
208
Pb/204Pb = 36.712 – 38.625 typical of ancient, high-grade terranes. Outcrops of Paleozoic
rocks of the Eastern Cordillera, some across the Western Cordillera and Altiplano-Puna (Fig.
1.3) have a lower metamorphic grade and contain more radiogenic 206Pb/204Pb > 18.5 than the
Arequipa basement (100 samples from Damm et al., 1990; Bock et al., 2000; Lucassen et al.,
2001). Jurassic rocks from the Coastal Cordillera (30 samples data from older igneous rocks
from Mukasa, 1986; Lucassen et al., 1996; 20 new samples) have 206Pb/204Pb=17.5-18.8. To
these and 30 published Rogers and Hawkesworth (1989), Haschke et al. (2002), 74 own new
data of Cretaceous igneous rocks were combined with 406 Pb-isotope ratios of Cenozoic
igneous rocks (70 samples from Trumbull et al., 1999; Kay et al., 1999; Siebel et al., 2001 and
336 own new data). Ore deposits (80 samples from GIS Andes BRGM; Macfarlane et al.,
1990, 1999; Puig, 1988; Tosdal, 1993; Vivallo et al., 1998). All data combined fall into two
groups: a low 206Pb/204Pb=17.5-18.5 in the central part (16°S-21°S) and high
206
Pb/204Pb=18.5-18.9 in the northern part of the Arequipa basement (16°S-13°S) and to the
south (21°S-28°S). The samples from the Eastern Cordillera (10 samples from Macfarlane,
1990; Aitcheson et al., 1995 and 10 new samples) have highest 206Pb/204Pb > 18.904,
207
Pb/204Pb > 15.667, 208Pb/204Pb > 38.944.
84
Fig. 4.1. a) Diagram of 206Pb/204Pb ratios versus Latitude °S in the Central Andes. Holocene volcanoes (HV),
Pleistocene volcanoes (PV), Mio-Pliocene volcanoes (MPV), Miocene volcanoes (MV), Paleocene-Eocene
volcanoes (PEV), Mio-Pliocene ignimbrites (MPIG), Miocene ignimbrites (MIG), Eocene-Paleocene-Cretaceous
intrusions (EPCIN), Jurassic intrusions (JIN), Back-arc volcanoes (BAV), Paleozoic basement (PB), Proterozoic
basement (PrB), Copper sulfurs (Cu), Silver-Gold sulfurs (Ag-Au) and ores along the back-arc (Ores-BA). Field
of Pacific mid-ocean ridge basalt are from Hergt and Hawkesworth, 1994. b) Histogram of Pb-isotope
compositions. Data sources of samples plotted here are explained in section 3.2 and 4.2.1.
4.3 Results
4.3.1 Mapping of Pb isotopes and geochemical variations in Pb-domains

The large number of Pb-isotope measurements of Andean rocks (Fig. 7.1) now available
allow to delineate crustal domains much more precisely than previously possible and to
construct present-day Pb isotope maps using GIS.
Spatial analysis of lead isotopes data established different domains and transitional zones,
which have special geochemical characteristics (Fig. 4.2 to 4.7):
The Arequipa domain presents lowest 206Pb/204Pb ratios (from 16.083 to 17.846), 207Pb/204Pb
ratios (15.435 to 15.612), 208Pb/204Pb ratios (35.625 to 38.491) in Proterozoic and Paleozoic
rocks. Slightly higher in 206Pb/204Pb from 17.846 to 18.551 ratios are found in Neogene arc
volcanoes and early Tertiary intrusions.
The Neogene volcanics in this domain have low εNd values (-5 to -9), Sr isotope between
0.7051 and 0.7073 and higher Sm/Yb (2 to 10) ratios.
85
The Transtion zones with 206Pb/204Pb = 18.551 to 18.727, 207Pb/204Pb = 15.612 to 15.652,
208
Pb/204Pb = 38.491 to 38.767 between the Arequipa and Cordillera domain.
The Neogene lavas in this domain have intermediate εNd values from -0.9 to -5, Sr isotope
ratios from 0.705 to 0.7065, high Sm/Yb ratios (2 to 10) in southern Peru, and low Sm/Yb
ratios (1.5 to 5) in northern Chile-southwestern Bolivia.
The Clemesi domain is defined by Mesozoic rocks of Coastal Cordillera. Lead ratios
206
Pb/204Pb= 18.7 to 18.4 of Mesozoic rocks are generally higher than the Proterozoic
basements (206Pb/204Pb = 16.7 to 18.4), and lead compositions 207Pb/204Pb, 208Pb/204Pb like the
basement on which they are located. Their higher εNd values (5 to -0.9) and low 87Sr/86Sr
ratios (0.703 – 0.705) are representative of a juvenile magma addition to the crust in Jurassic
and Cretaceous times. These rocks have low Sm/Yb ratios (< 3).
The Cordillera domain occurs to the south and north of Transition zones and has the highest
206
Pb/204Pb >18.727, 207Pb/204Pb > 15.652, 208Pb/204Pb > 38.767. Neogene volcanoes in this
domain have low εNd values (-0.9 to -9), Sr isotope between 0.7029 to 0.7051 and high Sm/Yb
ratios (5 to 9).
The Mejillonia domain with lead compositions 206Pb/204Pb = 18.375 to 18.551, 207Pb/204Pb =
15.573 to 15.652, 208Pb/204Pb = 38.170 to 38.767. The Jurassic rocks of la Negra Formation
have high εNd values (5 to -0.9), low 87Sr/86Sr ratios (0.703 – 0.705) and low Sm/Yb ratios
(1.5-2.5).
The Chilenia domain has similar 206Pb/204Pb, 207Pb/204Pb, 208Pb/204Pb ratios to the Mejillona
domain. Mejillonia and Chilenia domain are separated by a Transitional zone at 25°S.
Fig. 4.2. 206Pb/204Pb isotope map in the Central Andes (13°S -28°S).
86
Fig. 4.3. Map of present-day 206Pb/204Pb ratios and Fig. 4.4. Map of present-day 206Pb/204Pb ratios and
207
Pb/204Pb ratios of rocks from Central Andes. 208
Pb/204Pb ratios of rocks from Central Andes.
Fig. 4.5. Map of present-day 206Pb/204Pb ratios and εNd Fig. 4.6. Map of present-day 206Pb/204Pb ratios and
values of rocks from Central Andes. 87
Sr/86Sr isotope ratios of rocks from Central Andes.
87
Fig. 4.7. Diagram Sm/Yb ratios versus Latitude. Colors of the symbols are according to the Pb domains (see
illustration of Neogene volcanoes in Fig. 4.2.
4.4 Discussion
4.4.1 Pb isotope mapping crustal domains
The nature and timing of, and processes involved with, crust formation and the extraction of
the crust from the Earth's mantle remain subjects of much debate. The Pb isotopic system is
one of several isotopic systems that have been used to help understand the processes of crust
formation and has the advantage of being based on 3 parent-daughter systems with very
different decay rates and 2 elemental pairs (U/Pb and Th/Pb) with different geochemical
properties. Early studies of crustal Pb isotopic systematics (e.g. Oversby, 1978) established
that by the late Archean, crustal provinces existed with distinct Pb isotopic signatures.
Additional work has confirmed these distinctions and emphasized that igneous rocks inheret
geochemical characteristics from crustal sources, their isotope composition can “fingerprint”
basement terranes and provided useful information for delineated crustal domains/mapping
(Ayuso and Bevier 1991; MacFarlane et al., 1990; Aitcheson et al., 1995; Wörner et al., 1992)
and for evaluating plate reconstruction (Tosdal et al., 1996 and Loewy et al., 2004). Examples
are illustrated from the cordilleras of southwestern North America, Greenland, Antarctica,
Northern Andes and Central Andes.
In the western Unites States, Wooden et al. (1988), Wooden & DeWitt (1991) used present-
day Pb isotope compositions of Proterozoic rocks, and Mesozoic and Cenozoic plutonic,
volcanic rocks and ores to define distinct Proterozoic crustal-scale Pb provinces characterized
by different U-Th-Pb histories that broadly, but not precisely, correlate with known geologic
provinces (Fig. 4.8). Proterozoic Pb isotope provinces and Nd isotope provinces (DePaolo et
al., 1991) are similar in their geographic extent, but they are not the same. One area of
disagreement is the extent of the Mojave Pb province in Nevada, Utah, and Arizona. Given
that there are fundamental differences in the geochemical basis of the Pb and Nd isotope
systems, these differences in extent of crustal provinces may be real and/or rooted in sample
distribution and types. In contrast, the Pb isotope and Sr isotope provinces, particularly the
edge of significant Precambian crust marked by the initial Sr (Sri) equal to 0.706 (Fig. 4.8,
Kistler, 1990), correlate well (Wooden et al., 1998). Elison et al. (1990) showed that the Sri =
0.706 line also correlate with the shelf-slope break defined by Paleozoic and Triasic strata. Pb
isotopic mapping in Nevada, coupled with geophysical information (Rodriguez, 1998)
demonstrates that these boundaries are crustal scale faults.
88
Fig. 4.8. Pb isotope provinces in the western United States. From Wooden & DeWitt (1991). Sri = 0.706 line
from Elison et al., 1990.
4.4.1.2 Isotopic framework of basement in the Central Andes

The Central Andes is underlain by a complex geology (Fig. 3.1). However, unlike western
United States, Pb, Sr, Nd isotope frame work of the rocks were not well established because
an extensive cover by young volcanic rocks.
Macfarlane et al. (1990) used Pb isotope compositions of rocks and ore deposits of differing
ages over a broad geographic area to map out crustal provinces along and across the arc, and
they divided the Central Andean arc into three geological provinces (I, II, III) and six Pb
isotope subprovinces (Fig. 4.9a; Ia, Ib, Ic; II, IIIa, IIIb).
Aitcheson et al. (1995) documented a Pb subdivision in detail for the area in the Central
Andes from lat 16° to 24°S and from long 64° to 70°S and they differentiated in the northern
Altiplano a unradiogenic Pb isotope domain in ores and igneous rocks (Fig. 4.9b), and they
were not sure if this domain is part of the Arequipa massif. However, they noted that Nd
crustal extraction ages of ~1.9 Ga from northern Altiplano basement samples are similar to
the Protolith age inferred by Dalmayrac et al. (1977) for Arequipa massif granulites from U-
Pb zircon upper-intercept ages.
Tosdal (1996), based upon U-Pb geochronologic and Pb isotopic study, concluded that
Middle Proterozoic rocks in the Arequipa craton in western Bolivia and northern Chile are not
reworked Early Proterozoic rocks of Arequipa massif (Fig. 1.3). They are however, related by
similar unradiogenic 206Pb/204Pb < 18.2 reservoirs and be haved as a coherent block since the
Middle Proterozoic (Fig. 4.9c). Tosdal (1996) also proposed that parts of the Arequipa craton
are allochthonous with respect to Amazon craton.
Loewy et al. (2004) however, interpreted the Arequipa Antofalla craton as a single basement
block, which comprises three domains based mainly on geochronology: a northern domain
(2.02-1.79 Ga), a Central domain (1.2 – 0.94 Ga) and a southern domain dominated by
Ordovician metasediments derived partially from Proterozoic crust (Fig. 4.9e). Using
integrated data set that includes U-Pb geochronology, and the sequence of adjacent provinces
they suggested that the Arequipa Antofalla craton was accreted at ca. 1.05 Ga.
Lucassen et al. (2001) compiled from extensive geological and geochemical studies of
exposed pre-Mesozoic basement rocks in the Central Andes and presented an important
constraints on the compositional evolution of the crust in the Central Andes from 21° to 27°S
from the Early Paleozoic onwards (Fig. 4.9d). They also concluded that in terms of its Nd and
Sr isotope composition, the crust in this part of the Andes was already rather homogeneous at
the time Early Paleozoic metamorphism and granitic magmatism in the Paleozoic served to
increase this homogeneity. The geological and sedimentary record of the area indicates that
the crust that formed in the Early Paleozoic orogen was largely entire at the beginning of the
Cenozoic crustal thickening of the Central Andes.
89
b). Aitcheson et al., 1995
a). Macfarlane et al., 1990
d). Loewy et al., 2004
c). Tosdal, 1996 e). Lucassen et al., 2001
Fig. 4.9. Different lead provinces proposed by various authors in the Central Andes.
90
Based on tectonostratigraphic analysis of southern South America Ramos et al. (1988, 1986),
Bahlburg and Hervé (1997) presented a simplified terrane distribution maps. These terranes
define a mosaic of old continental crust (e.g. Arequipa-Antofalla, Mejillonia, Chañaral,
Chilenia, Pichidangui, Cuyania-Precordillera, Pampia terranes) amalgamated during the Late
Proterozoic to early Paleozoic times (Fig. 4.10a, b).
a) Ramos et al. (1988, 1986). b) Bahlburg and Hervé (1997).
Fig. 4.10. Maps of accreted terranes in the southern of South America.
In all these works mentioned above based on lead isotopes and tectonostratigraphic analysis,
and radiometric ages in the Central Andes is not well defined the real extension and locations
of terranes, mafic or felsic crust and transition zones between terranes. The crucial for the
reconstruction of the terrane assemblage of the Central Andes is the Arequipa terrane.
For the most part, boundaries for the Pb isotope provinces and terrane distribution presented
by Macfarlane et al. (1990), Aitcheson et al. (1995), Tosdal (1996), Loewy et al. (2004),
Lucassen et al. (2001), Ramos et al. (1988, 1986), Bahlburg and Hervé (1997) are similar in
their gross geographic location but differ significantly in their exact geographic position and
shapes. Differences in the exact locations of the boundaries in previous works are in function
of sampling density and interpretation.
The data set (Fig. 7.1) of this thesis is the most extensive used so far to delineate crustal
domain boundaries. Therefore the crustal boundaries are much better defined than in all
previous models.
Nd isotope analyses along these domains presented for the first time (Fig. 4.5, 4.12) further
support our crustal domain distinctions and corroborate the domain boundaries defined here.
Nevertheless, boundaries of crustal domains are not necessarily constrained by Sr isotopes
variations (Fig. 4.6) due to the effect of mixing of magmas, fractional crystallization, and
assimilation (see section 3.3).
High Sm/Yb (Fig. 4.7, 4.13, 4.14, 4.15) on thick crust implies a garnet residue, probably in
intermediate to deep crust in mafic to intermediate crustal rocks (Kay et al., 1999) as is the
case of younger volcanoes in the Arequipa domain and volcanoes in the Transition zone
between 16 and 14°S. Indeed most of the Proterozoic rocks of Arequipa terrane are mafic and
unradiogenic (Loewy et al., 2004; Aitcheson et al., 1995; and Tosdal, 1996).
91
Low Sm/Yb even on thick crust may imply either felsic crust as is the case of volcanoes in the
Cordillera domain between 22 and 28°S (Fig. 4.7) or assimilation occurred only in the upper
crust, which is unlikely. Lucassen et al. (2001) also agree that the crust in this region is felsic
(Fig. 4.9e and 4.12).
Fig. 4.11. Pb isotope domains and large-scale Fig. 4.12. Diagram of initial εNd and Sr isotope versus
structural trends in the Central Andes. The Contours to Pb isotope domains.
the Wadati-Benioff zone with depths in km.
Fig. 4.13. Diagram initial 87Sr/86Sr and εNd versus Sm/Yb of rocks from the different Pb domains. Legend as in
Fig. 4.12
92
Fig. 4.14. Crustal Pb-domain scale section (NW14°S – N18.5°S – S28°S). (after Wörner et al., 1992; Kay et al.,
1999).
Fig. 4.15. Crustal-scale section of Arequipa domain (NW17°S – NE15°S). Superimposed on crustal profile of
Allmendinger et al. (1997). Section line in Fig. 3.13.
The boundaries of these blocks reflect the positions and character of real geologic boundaries
between the different crustal blocks and these follow large-scale structural trends (Fig. 4.11).
The abrupt boundary between Arequipa domain and Transition zone at ~16°S coincides with
the Iquipi fault and with the Colca Canyon. Toward the East between the Transition zone and
Cordillera domain could be a major fault parallel to the Andahuayllas-Copacabana fault
system. The transition of Arequipa domain to the east may be related to the Copacabana-
Coniri fault system. The boundaries to the west of Arequipa domain and the east boundary of
Clemesi domain coincide with Incapuquio fault system. The transition zone between the
Cordillera domain and Mejillonia – Chilenia domains coincides with West fissure fault
system in northern Chile.
4.4.2 Crustal Contamination in the Pb domains

It is mostly accepted for Central Andean Miocene to Recent magmas that Nd-Sr isotope
mostly variations reflect the process and degree of crustal contamination (Davidson et al.,
1992). There is a general pattern of decreasing εNd and increasing 87Sr/86Sr with time for rocks
of each Pb domain. Jurassic rocks and early Tertiary intrusions are less contaminated than
Neogene volcanoes.
The lowest εNd values and low 87Sr/86Sr ratios have the Neogene volcanoes in Arequipa
domain. The low εNd values and high 87Sr/86Sr ratios have the volcanoes of the Transition
zones and Cordillera domain.
93
Figure 4.16 shows a simple mixing model using the parameters of the Table 4.1. Mixing
curve for Arequipa domain between Cretaceous granite (67 Ma) and Proterozoic gneiss (2
Ga), and for Cordillera domains in northwestern Argentina between Miocene andesite (19.6
Ma) and Paleozoic orthogneiss (460 Ma).
Table 4.1. Values used in the mixing model. Data are from a-Boily et al. (1989), b-James (1982), c-Thrumbull et
al. (1999), d-Lucassen et al. (1999b).
87
Samples (Long.°W, Lat.°S) Age (Ma) Sr Nd Sr/86Sr initial 144
Nd/143Nd
Arequipa domain (70.70, 17.29)a 67 401 24.4 0.7049 0.512739
Arequipa domain (71.5, 16.5)b 2000 150 36.0 0.720 0.511500
Cordillera domain (63.38, 25.9)c 19.6 414 21.9 0.7050 0.512634
Cordillerra domain (57.5, 24.6)d 460 143 32 0.7235 0.511810
The mixing model result are two curves, which show that the smallest amonts of crustal
contamination in Early Tertiary intrusions (≤ 10%). In Arequipa domain Miocene lavas have
≤ 25% and the maximum crustal contamination (40 %) is modeled for the Mio-Pliocene to
Recent lavas. In Cordillera domain Miocene lavas have ≤ 15% and the Mio-Pliocene to
Recent lavas have 40% (Fig. 4.14).
The simple mixing curve (dashed line) for ignimbrites in the Cordillera domain and
Transitional zone shows that (87Sr/86Sri)/εNd ratios of their magmas are lower than
(87Sr/86Sri)/εNd ratios of crust.
Fig. 4.16. Diagram of εNd versus 87Sr/86SrInitial for rocks of the Central Andes. Circles mark percentages of crustal
contamination of specific domain. NVZ=north volcanic zone (Bourdon et al., 2002), SVZ=south volcanic zone
(Kay et al., 2005) and AVZ=austral volcanic zone (Stern and Killian 1996). Legend as in Fig. 4.12.
It have presented a simple mixing model because is a regional analysis, advances models of
open system magmatic processes like AFC (Assimilation Fractional Crystallization: after
DePaolo, 1981; O’Hara and Mathews, 1981), EC-AFC (Energy-Constrained Assimilation
Fractional Crystallization: Spera and Bohrson, 2001), MASH (Mixing Assimilation Storage
Homogenization; Hildreth and Moorbath, 1988), RAFT (Recharge, Assimilation,
94
Fractionation, Tapping: Aitcheson and Forrest, 1994) need estimate values of local parameters
such as F (mass of remaining magma, mass of original magma) and r (rate of assimilation/rate
of fractional crystallization), O-isotopes ratios, Tm (magma temperatures), Ta (assimilant
temperature). These models are more useful in individual volcanoes or local areas and
previous studies of volcanic complex have calculated for Parinacota volcano (Bourdon et al.,
2000), Taapaca volcano (Kohlbach, 1999), El Misti volcano (Kiebala in preparation).
However, we cannot just explain by fractional crystallization the large variability in trace
elements of volcanic in the Central Andes.
4.4.3 Isotopic domains constrained by 3D density model

The three-dimensional density model of Tassara et al. (2006) was designed to represent the
current distribution of mass along the Andean margin (5°-45°S) at continental and
lithospheric scales. The structure of the model is formed by a number of bodies simulating the
subducted slab, the sub-continental mantle and the continental crust. Each body has one value
of density that was selected in accordance to its position into the model and after studying the
dependency of this physical parameter on chemical composition, metamorphic pressure-
temperature conditions, water content and degree of partial melting. To reduce the degrees of
freedom during the forward modelling of the Bouguer anomaly, the geometries of the slab,
lithosphere-asthenosphere boundary, and continental Moho were fixed, as far as available
geophysical data allow (for location and sources of data see Tassara et al., 2006). The existing
geophysical database for the Central Andes constrains the subcrustal and Moho geometry.
The one remaining degree of freedom is the geometry of intracrustal density discontinuity
(ICD), separating an upper-crustal body of density 2.7 g/cm3 from a lower-crustal body of
density 3.1 g/cm3. Following empirical relationships between density and silica content of
crystalline, anhydrous crustal rocks (Tassara, 2006), the former simulates a granitic upper
crust of ca. 70 wt% SiO2, whereas the dense lower-crustal body represents a garnet-pyroxene
granulite of 55-58 wt% SiO2. If the lower crust below the Central Andean orogen is accepted
to be partially hydrated and containing some degrees of molten material, then the selected
bulk density of 3.1 g/cm3 represents even more basic compositions (55-48 wt% SiO2
following Tassara, 2006 ).
The geometry of the ICD is a proxy to regional-scale (several tens to some hundred
kilometers) lateral density variations occurring inside the crust. The main factor controlling
such variations is the bulk compositional structure of the crust, i.e. the vertically-integrated
proportion of felsic to mafic crust, whereas spatial variations of lower crustal temperature,
hydration and partial melting could have a secondary effect (Tassara, 2006; Tassara et al.,
2006).
Using the final geometries of the 3D density model in their electronic version (Tassara et al.,
2006), we computed a grid covering the study area for what we call the “crustal structure
index Ө”, i.e. the ratio between the thickness of the upper-crustal body and the total crustal
thickness. Low (high) values of Ө indicate a predominance of mafic (felsic) material in the
crust.
For this study we combined the result of crustal structure index (Ө) mapping and our Pb
isotope analysis, and they show a good correlation (Fig. 4.17):
¾ Low Ө < 0.2 coincides with Pb isotopes of Arequipa domain.
¾ High Ө > 0.2 correlates with Pb isotopes of Transitional zones and Cordillera domain.
This correlation between 3D density model and Pb domains indicate that the age and
proportion between felsic and mafic crust are the factors controlling the density structure
inside the crust and isotopic and geochemical variations along the Central Andes.
The highest Sm/Yb ratios (2 to 10) in the youngest volcanoes (< 3 Ma) typically occur on
more mafic crust (Arequipa domain) and the Transition zone in southern Peru. On more felsic
95
crust (Cordillera domain in northern Chile and northwestern Argetina) Sm/Yb ratios (2 to 6)
increase with time but never reach these high levels obtained in Arequipa domain. This
confirms that bulk crustal composition is another control of the garnet signature in addition to
crustal thickness. Petrophysical modelling (i.e. Sobolev and Babyeko, 1994; Tassara, 2006)
demonstrated that mafic rocks are able to form more garnet than felsic rocks for the same
pressure-temperature conditions (i.e. crustal depth), explaining why the garnet-effect (high
Sm/Yb ratios) is apparently stronger in the Arequipa domain compared to the Cordillera
Domains.
Fig. 4.17. Map of crustal structure index Ө, (Tassara et al., 2006). Circles are locations of the different
206
Pb/204Pb ratios.
4.4.4 The Nature of the Transition Zones

The Transitional Zone to the N of the Arequipa domain was traversed by magmas, which
have high Sm/Yb ratios indistinguishable from those of the Arequipa domain, yet their Pb
(and Nd) isotopic character is distinctly different for all young volcanic rocks. Thus, the
boundary between the Arequipa Domain and the northern Transition Zone should be abrupt
and possibly near-vertical within the crust, because the Pb-signature is the same (within a
domain) for mafic and silicic rocks that probably sample preferentially the lower and upper
crust. This suggests that the boundary may be a deep transversal crustal fault (e.g. Iquipi fault,
Fig. 4.11), unlike the southern boundary of the Arequipa Domain (Wörner et al., 1992). In
96
addition, a high-P garnet signature is present in the northern Transition Zone near the
boundary even where the crustal structure index Ө is high, indicating the absence of a thick
mafic lower crustal layer. This apparent contradiction to the general pattern predicted above
may be due to the poor lateral resolution of the crustal structure index Ө and the poorly-
known physical nature of the crustal boundary.
From 21 to 22°S is located the Transitional zone between Arequipa and Cordillera domain
(Fig. 4.18). Seismic studies by Zandt et al. (1996), Swenson et al. (2000) beneath this area
have concluded that the average velocity of the crust is low (Vp = 5.9 to 6.2 km/s, see e.g.
Zandt et al., 1996), and its average Poisson’s ratio is about 0.25 (Vp/Vs = 1.73) and this was
interpreted as an evidence for the felsic composition of the crust.
However, Yuan et al. (2002) showed that the average correlation between topography and
Moho depth changes at 21° to 22°S. They suggest that in this region the crust contains a thick
lower crustal mafic component below the Altiplano. They suggested a less proportion of
mafic crust further to the S as a consequence of the delamination and detachment of high-
grade mafic metamorphic lower crust. Lower crustal delamination below the Puna plateau
was also suggested by Kay et al. 1994. Recent studies (e.g. Lucassen et al., 2001; Swenson et
al., 2000) confirm that the lower crust below the Puna region may be felsic- rather than mafic-
dominated.
Our geochemical and isotopic data for igneous rocks between 21° and 22°S (Fig. 4.3, 4.4, 4.5,
4.6 see also Wörner et al., 1992) also suggests the presence of mafic crust at depth beneath
this area. Apparently the southern margin of the Arequipa mafic terrane extends toward 22°S
and is covered the Phanerozoic felsic rocks (Fig. 4.14)
4.4.5 Segmentation of the Central Andes plateau related to the Pb-isotope domains
In the last decades several authors have discussed the role of inherited pre-Andean crustal
structures in the upper plate on the Cenozoic evolution and segmentation of the Central Andes
(Allmendiger and Gubbels, 1996). Jordan (1988) suggested that the Altiplano formed across
the western margin of a thick and stabilized cratonic lithosphere or “Tectosphere” and that the
thick Paleozoic sedimentary sequences may have localized thin-skinned tectonics during
shortening (Sclater et al., 1980; McQuarrie and DeCelles, 2001). The Puna segment, its
morphological characteristics and the thick-skinned tectonics of the Sierras Pampeans was
proposed to have developed over a thermally thinned continental lithosphere and that this
segmentation is controlled by the internal crustal structure and delamination in this segment
(Kay et al., 1994 and reference therein).
The crustal domain boundaries newly defined here should help to further constrain the role of
pre-existing crustal heterogeneities in the evolution and deformation pattern of the Central
Andes:
Figure 4.18 shows that the segmentation of the Central Andean plateau and its adjacent
foreland is in fact spatially related to the Pb-isotope domains: The Arequipa domain is
coincident with the broad high Altiplano plateau (3.8 km high and 300 km wide). The largest
amount of shortening in the Eastern Cordillera and Subandean belt during Andean Orogeny is
located to the E and NE of the Arequipa domain. The Transition Zone (between 21°S and
22°S) coincides with the transition between the Altiplano and Puna segments. In the region of
the southern Cordillera Domain, the Puna plateau is higher (4.4 km) than the Altiplano and is
much narrower. The foreland structure is dominated by thick-skinned Laramide-style
basement uplift on steeply E- and W- dipping reverse faults (Santa Barbara System;
Allmendiger and Gubbels, 1996). Therefore I argue that the nature, i.e. the bulk composition
and thus the different rheologies of the crust are an important factor that controlled the
deformation pattern of the Central Andes and the localization of the Andean plateau.
97
4.4.6 Rheological and structural identity of the Pb domains during Andean Orogeny
Crustal architecture and evolution are also different with respect to the sedimentary cover:
The thickness of Paleozoic to Cenozoic sediments is small, if at all present in the Arequipa
domain. Outcrops of metamorphic basement occur below a relatively thin cover of Jurassic
and Tertiary sediments. By contrast, significant thicknesses of older sediments, which are
strongly deformed by the Andean orogeny, are observed predominantly in areas surrounding
the Arequipa domain. Conversely, erosion products deposited during uplift of the Central
Andes (e.g. Moquegua and Azapa Formations and equivalents, e.g. Semperé, 2000, Roperch
et al., 2006) appear to be much thicker on the Western Andean Escarpment (Fig. 3.1 and Fig.
4.18) in the area of the Arequipa domain. Furthermore paleomagnetic study combined with
stratigraphic analysis of Neogene basin on the Altiplano plateau (e.g. Rousse et al., 2005)
suggested that the Huaccochullo and Corque basins (Fig. 4.18) are deformed and rotated more
or less coherently, while the region directly east (e.g., the Copacabana-Coniri-Cochabamba
fault system: Fig. 4.11) accommodated the bulk of the shear component associated with the
large block rotation. These differences attest to the rheological and structural identity of the
Arequipa domain during Andean Orogeny.
Fig. 4.18. Central Andean plateau and its adjacent foreland related to the Pb-isotope domains. Extensions of
Huaccochullo and Corque basins from Rousse et al. (2005), Moquegua-Azapa basin from Roperch et al. (2006)
and Inner Arc from Clark et al. (1990).
98
4.4.7 Pb-isotopes of gold, silver copper and tin deposits, their relationship with the Pb-
domains
Lead isotopes tell us about the influence of basement rocks and tectonic setting on (Pb-)
sources of ore deposits in magmatic arcs. Thus, the application of crustal scale Pb isotope
variation contributes to an understanding of regional controls on ore deposits (Tosdal et al.,
1999).
The geology and geochemistry of the mayor ore districts in the Central Andes have been
described by Tilton and Barreiro (1980), Barreiro and Clark (1984), Puig (1988), Boric et al.
(1990), Macfarlane et al. (1990), Mukasa et al. (1990), Vivallo et al. (1998), Clark et al.
(1990a and b), Tosdal (1995), Richard et al. (2001), Kamenov et al. (2002), and BRGM GIS
Andes Project (see http://gisandes.brgm.fr).
Economic copper deposits are hosted in Jurassic and early Tertiary rocks in the Arequipa,
Mejillonia, and Chilenia domains and they have transitional Pb isotope compositions. Silver-
gold deposits are hosted in a variety of rocks across several Neogene volcanoes in
Transitional zone and Cordillera domain boundary with Pb isotope ratios comparable to that
of the Paleozoic basement (Fig. 4.1, 4.19).
As is observed in Figure 4.20, there are no abrupt Pb-isotope distinctions among Jurrassic (La
Negra Formation; 206Pb/204Pb = 18.1-18.5) and Early Tertiary igneous rocks (Toquepala and
Cerro Verde complexes; 206Pb/204Pb= 18.3-18.7) that host the copper deposits. However, there
are Pb isotope distinctions between Pb-isotopes of these copper deposits and Neogene silver-
gold deposits with 206Pb/204Pb > 18.5. This implies that the Cu ores and Ag-Au ores may have
significantly different sources.
Lead isotope compositions (206Pb/204Pb > 18.5) of Neogene volcanoes that host the silver-gold
deposits form trends similar of those of the local Paleozoic basement rocks. Lead isotope
compositions of Jurassic and Early Tertiary rocks (206Pb/204Pb= 18.7-18.4) generally have
higher ratios than the Proterozoic basements (206Pb/204Pb = 16.7-18.4) on which or near to
which they are located (Fig. 4.1). This is also true from Cenozoic volcanoes emplaced on
Proterozoic rocks.
This difference is in part the result of the time integrated growth of Pb isotope compositions
in response to widely different U/Pb and Th/U in Proterozoic rocks.
Jurassic and Early Tertiary tectonics in southern Peru and northern Chile were dominated by
large-scale faults, extension and crustal thinning that was accompanied by wide-spread
tholeitic and cal-alkaline magmatism (Mukasa et al., 1990; Lucassen and Thirlwall, 1998). In
contrast, Neogene volcanic were emplaced during and after crustal thickening (James and
Sacks, 1999; Kay et al., 1999).
Jurassic, Early Tertiary rocks and sulfide minerals of Arequipa, Mejillonia domains are also
isotopically distinct from Neogene rocks and sulfide minerals of Transition zone and
Cordillera domain (Fig. 4.21). The Jurassic and Early Tertiary sulfide minerals in porphyry
deposits are lower in 206Pb/204Pb < 18.7 (Vivallo et al, 1998; Tilton and Barreiro, 1980; Puig,
1988) as most of the igneous rocks with which the sulfides are associated.
In the Precordillera arc (northern Chile) Late Cretaceous–Eocene (68 Ma, Wörner et al., 2000
to 38.5 Ma, Scheuber et al., 1994) alkalic magmatism occurred in an extensional tectonic arc
and backarc setting (Scheuber et al., 1994, Charrier and Reutter, 1994), the Pb isotope
compositions (206Pb/204Pb =18.65) of these rocks are distinct from the Jurassic plutons and the
difference reflect extensional tectonic setting and derivation from the mantle source without
crustal assimilation (Fig. 4.21).
Of note is the fact that Pb isotope compositions of other some Jurassic rocks of the Coastal
Cordillera are devoid of porphyry copper deposits (e.g., Morro de Arica with 206Pb/204Pb=
99
206
18.8 and Punta Coles with Pb/204Pb= 17.58) and these rocks were formed with less crustal
interaction.
Fig. 4.19. Distribution of mineral deposits and Pb isotope domains.
Fig. 4.20. 206Pb/204Pb of Proterozoic, Paleozoic, Jurassic, Early Tertiary and Neogene crustal rocks versus
206
Pb/204Pb of sulfide minerals from different domains.
100
Fig. 4.21. a) 208Pb/204Pb versus 206Pb/204Pb, b) 207Pb/204Pb versus 206Pb/204Pb diagrams for Proterozoic, Paleozoic,
Jurassic, Early Tertiary rocks and sulfide minerals. Fields with dashed lines represent Pb isotope values of
Neogene volcanoes in differents Pb domains.
Crustal interaction is important for Jurassic and Early Tertiary porphyry-copper deposits
(Santo Domingo, Mantos Blancos, Michilla, Cerro Verde, Toquepala) and principally for
Neogene epithermal-silver-gold deposits (Shila, Colquechaquita, Potosi, Caylloma,
Magallanes, Todos Santos, Orcopampa).
The gradual increase of Pb isotope ratios with time within productive Jurassic and Early
Tertiary igneous rocks and ore minerals in the Arequipa domain suggest progressively
involvement of Proterozoic Arequipa crust material with lower U/Pb and a less assimilation of
high U/Pb crust in magma genesis. A scenario where mantle-derived magmas underplate,
assimilate lower mafic crust with lower U/Pb and in the upper part is added high U/Pb crust
seems to best explain the evolution of Pb isotope composition from Jurassic to Early Tertiary
times (Fig. 4.22).
Fig. 4.22. Model of Jurassic and Early Tertiary rocks and porphyry copper formation in the Arequipa domain.
101
Neogene crustal thickening accompanied by emplacement of magmas leads to a progressively

warmer crust, consequent lessening of the density contrast between those magmas and crust,
and stalling the magmas in the lower crust. Ponding magma further heat the crust, causing
melting of lower crust rocks and either the rise of a separate magma or the assimilation of
these partial melting into the ponded magma (Babeyko and Sobolev, 2002). Additional partial
melting, fractionation, assimilation of thick Paleozoic and Meso-Cenozoic crust with high
U/Pb can modify the chemical characteristic of the magma prior to or at site of emplacement
(Fig. 4.23). This is also constrained by Nd-Sr isotope of the volcanoes in the Transition zones
(Fig. 4.16). I propose thicker crust during Miocene time may have promoted ponding magmas
at upper levels in the crust and extensive assimilation of felsic rocks (section 4.4.2).
Fig. 4.23. Model of Neogene and Epithermal silver and gold formation in the Transition zone.
The Pb isotope results of this study concur with Kamenov et al. (2002) study and their
identification to the east of the Arequipa basement (Fig. 4.24a) and in the notion that the
degree of incorporation of ore metals from the basements appears to depend on the timing
and/or location of the mineralization event. My domain map defines the margins of the
Arequipa domain more clearly (Fig. 4.2).
I disagree with Chiaradia and Fontbote (2002) in the notion that mantle heterogeneity is an
important factor responsible for Pb isotope variability of the Andean provinces at the
continental scale (Fig. 4.24b). This may be true for the Northern Andes and thinner crust, but
in the Central Andes the heterogeneity of the crust (mafic and felsic) dominates.
102
a) Kamenov et al., (2002)
b) Chiaradia and Fontbote. (2002)
Fig. 4.24. a. Geographic distributions of ore lead isotope provinces in the Central Andes (Kamenov et al., 2002).
b. Map of Western South America showing the four lead isotope provinces of the Andean Cordillera defined by
Chiaradia and Fontbote. (2002), 1=Ecuador–Colombia, 2 =Northern Peru, 3= Southern Peru, 4=Chile.
Different to the Copper, Silver and Gold deposits, the Tin deposits are restricted to the central
section between 14°S and 22°S (Fig. 4.19). This belt, referred to as the Inner arc (Clark et al.,
1990), is restricted to the Eastern Cordillera of the centermost Andes, and lies to the east of
the zone of high Altiplano plateau (Fig. 4.18).
Fig. 4.25 Fig. 4.26
Fig. 4.25. The inner arc with tin belt metallogeny in the Central Andes (from Mlynarczyk and Williams-Jones,
2005). (a) Migration of mineralized centers with time (b) Permo-Triassic, early Jurassic and Tertiary plutons and
volcanics in the Central Andean tin belt.
Fig. 4.26. Schematic cross-section of the Andean continental margin, showing the relationships between the rate
of convergence, the corresponding retreat of the subducted slab, the angle of subduction, and the width of the
magmatic arc (from Mlynarczyk and Williams-Jones, 2005).
Studies by Mlynarczyk and Williams-Jones (2005, Fig. 4.25, Fig. 4.26) concluded that
Tertiary magmatism in the Inner arc of the Central Andes and the associated world-class tin
mineralization are directly related to discrete episodes of “collisional” interaction between the
Nazca slab and the South American plate. During these geodynamic events, the
asthenospheric mantle wedge, undergoing intense melting, was shifted eastward by a
103
shallowing of subduction and caused the underplating of thick continental crust by hot mafic
melts. This, together with shear heating generated by crustal thrusting, radiogenic heat from
crustal thickening, and the depression of the lower crust into the mantle caused an extensive
anatexis of the thick sequences of metasedimentary rocks and the gneissic basement, which
underlie the Inner Arc. Voluminous, tin-rich magmas of peraluminous composition were
produced and upon emplacement gave rise to large deposits of high grade tin mineralization.
They infer that this process has taken place repeatedly since the onset of the Andean orogeny
and, therefore, concluded that it may also be responsible for the earlier episodes of tin
mineralization in the Central Andes.
If it is as Mlynarczyk and Williams-Jones (2005) concluded why the tin beld don’t extent
along all “colisional” zone e.g. south of 22°S under the Puna plateau and north of 14°S under
the Altiplano plateau?
I observe that the Tin deposits and the high heat flow region of the inner arc coincide spatially
with the eastern bondary of Arequipa domain (Fig. 4.18, 4.19). Thus appear that the Tin beld
are related to the eastern Transition zone between the Arequipa domain and Cordillera domain
(see Fig. 4.15).
4.5 Conclusions
¾ The trace elements and Sr-Nd-Pb isotope ratios of Cenozoic lavas and ignimbrites
formed during and after crustal uplift follow patterns consistent with their basement:
volcanics erupted through Arequipa domain are affected by assimilation of mafic
basement. Lavas and ores on Mejillonia and Chilenia domains have received mafic
crust contribution. Volcanics in northern and southern Transitional zones were
contaminated by less mafic basement and more felsic basement. Lavas from Cordillera
domain are affected by assimilation of felsic basements. Clemesi domain contain more
juvenile component.
¾ Caution needs to be applied when plotting geochemical data vs. age for rocks from a
wide regional distribution as local basement control is more important than age in
controlling their geochemical composition.
¾ The correlation between the present-day Pb-isotope domains and crustal structure
index variations strongly supports an intracrustal control of the lead isotope
compositions of igneous rocks and ore deposits. The main control on Pb isotope
compositions is clearly the time-integrated trace element compositions of mafic and
felsic basement.
¾ Bulk crustal composition is another control of the garnet signature in addition to
crustal thickness. Garnet-effect (high Sm/Yb ratios) is higher in the mafic domains and
low in felsic domains.
¾ The relative abundance of ore deposits that fall into boundary of the mafic domains
with which they share their Pb-isotope compositions suggest that this crustal structure
and composition (as expressed also in the index model) are may be more favorable to
the formation of porphyry copper deposits. By contrast, The Transitional zones and
The Cordillerra domain (felsic crust) are may be more favorable to the formations of
epithermal silver-gold deposits, and the east Transitional zone between Arequipa and
Cordillera domains are may be favorable to the formation of tin deposits.
¾ On the basis of the geochemical and isotopic parameters discussed in this chapter and
geophysical model, the crustal rheology is peculiar in each basement domain.
104
5 Appendix
5.1 Sample locations

Data of 1545 samples rock analysis from the Central Andes were used in this study (Fig. 7.1).
The vast majority of the samples represent lavas and ignimbrites from the Miocene, Mio-
Pliocene, Pleistocene and Holocene arcs rocks. Additional samples are Proterozoic and
Paleozoic metamorphics rocks, Jurassic, Cretaceous and Paleogene intrusions, and ores. For
better graphical display, the samples are grouped in 18 groups (Table 7.1 and on CD).
Table 7.1. Table of age groups and numbers of samples.
Mamani Wörner et
al. and Literature Total of
Frontal arc Age (Ma) work unpublished samples samples
samples samples
HV= Holocene volcanoes 0 - 0.3 46 380 8 434
PV= Pleistocene volcanoes 0.8 - 3 26 19 2 47
MPV= Mio-Pliocene volcanoes 3 - 15 67 39 10 116
MV= Miocene volcanoes 15 - 25 12 14 18 44
EOV= Eocene-Oligocene volcanoes > 25 9 9
MPIG= Miocene to Recent ignimbrites 0 - 15 149 38 187
MI= Miocene Ignimbrites 15 - 25 17 6 23
Back Arc
HVBA=Holocene volcanoes > 0.7 6 7 13
MPVBA=Miocene-Pliocene volcanics 3 - 15 72 72
OVBA=Oligocene volcanics 28 - 30 28 29 57
MPIGBA=Mio-Pliocene Ignimbrites 5 - 21 17 17
MINBA=Miocene intrusions ~ 13 -15 1 8 9
INBA=Oligocene Intrusions 28 - 30 17 17
Intrusion
CPIN=Cretaceous-Paleocene intrusions 90 - 33 8 80 80 168
JIN=Jurassic intrusions 100 - 190 7 3 37 47
Basement
PB=Paleozoic 502 - 298 4 21 90 115
PR=Proterozoic 561-2000 8 11 24 43
Ores 8 - 120 7 120 127
Description of the volcanoes from the CVZ listed in the database can find on a website listed
in Table 7.2.
Table 7.2. Volcanoes Website.

Website
Peru volcanoes http://www.igp.gob.pe/vulcanologia/
Central Volcanic Zone volcanoes http://volcano.space.edu/cvz/volcindx.html
Globalvolcanism (Holocene volcanoes) http://www.volcano.si.edu/world/allvolcs.htm
The Figure 7.1 shows all Andes Project samples, which have been taken during 1988-2004,
while the Figure 7.2 on CD is located only those samples with complete analyses (XRF-
ICPMS-TIMS). Samples with just XRF and ICPMS data from the campaigns 1999 to 2004
are not illustrated on the Fig. 7.2.
105
Fig. 7.1. Samples locations.
5.2 Analytical Methods

In this section I describe the analytical methods used in this work and in Wörner et al. after
1993 and unpublished data, which have been done at the Geowissenschaftliches Zentrum
Göttingen (GZG).
Samples from Wörner et al. before 1993 were analyzed by Atomic Absorption Spectrometry
and Instrumental Activation Analysis. Some samples of this group were analyzed again by
ICP-MS in this study in order to obtain a complete set of HFSE and REE.
5.2.1 Sample Preparation

Representative samples from CVZ were collected, analyzed by XRF and ICP-MS and TIMS.
The most unaltered rocks were selected in the field to minimize contamination of the samples,
removal of alteration rims and crushing up into chips of ~5 cm in diameter was done directly
at the sampling location. About 1-3 kg of each sample was collected and a hand-specimen
was kept from all samples.
The chips were then disintegrated with the small jaw crusher down to fragments of less than 1
cm in diameter. About 100 g of samples material was separated from the remaining fragments
106
with a mechanical splitting device. This 100 g were then ground to powder (65 μm) in agate
mills for about 30 and 40 min. The rest of the samples not used to ground, some of which
were used to mineral separation.
5.2.2 XRF (X-Ray Fluorescence Spectroscopy) analysis

XRF spectroscopy uses characteristic x-rays (secondary fluorescence), generated by
excitation of inner shell electrons with primary X-Rays to specify elements and concentration.
This analytical method was used to determine major- and minor elements (Si, Ti, Al, Fe, Mn,
Mg, Ca, Na, K, P) of the rocks as well as some trace elements (Nb, Zr, Y, Sr, Rb, Ga, Zn, Cu,
Ni, Co, Cr, V, Ba, Sc). 700 mg of powder sample were thoroughly mixed with 4200 mg
Spectroflux 100 (Dilithiumtetraborate [Li2B4O7]) and automatically fused and transformed to
a glass disc by an automatic fusion devise (Labor-Schoeps SGE 20). Fe2O3 was determined
titrimetrically with KMnO4 and the loss on ignition (LOI) by weight difference at heating to
1100°C. Analytical error for major elements are around 1% (except for Fe and Na, 2%; and
LOI, ~10%) and for trace elements around 5%. Measurements were carried out on a Philips
PW 1480 and the data processing is controlled by the Philips X40 software package. For the
calibration of major and trace element determination were used about 50 reference materials:
a wide variety of international geochemical reference samples from the US Geological
Survey, the International Working Group "Analytical standards of minerals, ores and rocks",
the National Research Council of Canada, the Geological Survey of Japan, the South African
Bureau of Standards, the National Institute of Standards & Technology, etc.
5.2.3 ICP-MS (Inductively Coupled Plasma Mass Spectrometry) analysis

The ICPMS technique provides the advantage of simultaneous determination of a wide range
of trace elements from Li to U, low detection limits and good analytical precision especially
for masses >80. Fundamentals of this method are described in Potts (1987), Garbe-Schönberg
(1993). For the present purpose, trace elements like REE (La – Lu), HFSE (Nb, Ta, Zr, Hf),
LILE (Rb, Sr, Ba, Cs), and others (Li, Sc, Cu, V, Cr, Co, Ni, Zn, Y, Pb, Th, U) of selected
samples were analyzed on an ICPMS Perkin Elmer SCIEX (Elan DRC II).
Sample preparation to ICPMS: About 100 mg of sample powder was weighed into Teflon
beakers after Heinrichs & Herrmann (1990) and dissolved under pressure in a mixture of
acids composed of 3 ml HF, 3 ml HClO4, and 1 ml HNO3. After cessation of gas formation,
the beakers were tightly closed and heated in a muffle oven at 200°C for 14 hours. After
cooling to ambient temperatures, the beakers were opened and placed on a hot plate for
evaporation at 180°C. The dried sample was re-dissolved in 2 ml HNO3 and once again put
onto the hot plate. After the second stage of evaporation, 2 ml HNO3, 5 ml de-ionised water
and the internal standard were added to the sample. This solution was rinsed into a 100 ml
quartz-glass flask and the remaining volume was filled with de-ionised water. The resulting
solution contained the sample in a dilution 1:1000 in 2 % HNO3 with 20 ppb Rh, Re, In as
internal standard. It was kept in polyethylene bottles and measured within two days to
minimize the risk of precipitation or interchange with the plastic container. For each batch of
fifteen samples a blank solution and a lab-intern or international standard were prepared. All
reagents used were three times distilled to suppress blank values.
Selection of respective isotopes was done on criteria of abundance and exclusion of isobaric
interference. Calibration of the intensities against concentration values for each element is
provided by measuring multi-element calibration solutions prior to sample acquisition. These
calibration solutions were prepared from stock solutions by aiming to span the total range of
expected concentrations and by the attempt to keep the elemental ratios close to natural
conditions. A control for contaminations due to dissolution processes is obtained by
measurement of simultaneously prepared blank solutions, a correction for the machine drift is
107
provided by repeated measurement of one of the calibration solutions and by the internal
standards of Rh, Re, In. Co-processed lab-internal and international standards JB3 (n= 19) and
JA2 (n = 8) were measured as unknowns in the experiments. From this the 2σ error of the
method is estimated to be <20% for Nb and Ta, <10% for Be, Cs, Cu, Hf, Li, Y, Pb, Rb, Tl,
Th and U and ~5% for the rare earth elements (REE).
5.2.4 TIMS (Thermal Ionization Mass Spectrometer) analysis

The isotope ratios of Sr, Nd and Pb on corresponding 194 whole rocks were measured at a
Finnigan MAT262-RPQII plus in Göttingen (Abteilung Radiogene Isotope).
For Sr and Nd isotopic determination 100 mg of sample powder was dissolved in 6 ml HF:
HNO3 (1:1) for 16 hours at 200°C baked within Savillex vials. The solution was evaporated
to complete dryness at 140°C on a hot plate, dissolved and evaporated two times again in 4 ml
6N HCl , and evaporation for the last time, it was re-dissolved in 2.5 ml 2.6 N HCl, stored in
PE vials and centrifuged. For separation the sample solution was rinsed with 2.6 N HCl
through columns containing ion exchange resin BIORAD AG 50W-X8 Resin, 200-400 mesh.
The strontium rich elution fraction was caught in a vial, evaporated to dryness and stored until
measuring. For Nd separation, the REE rich fraction gained from the above separation
sequence was separated in a second set of columns containing Teflon powder which is
impregnated with ion-exchanging HDEHP Bis-(2-etylhexy)-Phosphate. Elution of Nd was
done with 0.18 N HCl.
For measurement, Sr was dissolved in 0.5 N H3PO4 and mounted on Re-double filaments (~1
μg), and Nd was dissolved in 2 N HCl and mounted on Re-double-filaments (~1 μg). The Sr
and Nd isotope ratios were corrected for mass fractionation to 87Sr/86Sr = 0.1194 and
143
Nd/144Nd = 0.7219 and normalized to values for NBS987 (0.710245), and La Jolla
(0.511847), respectively. Measured values of these standards over the period of the study
were 0.710262 ± 24 (21 analyses) and 0.511847 ± 20 (12 analysis). External error 2σ are
estimated at < 0.004% for Sr and Nd. Total procedural blank: Sr (0.26 ng) and Nd (< 0.14 ng).
For Pb isotopic determination about 100 mg of sample powder was dissolved in 4 ml HF:
HNO3 (1:1) in Savilex beakers at 200°C for 24 hours. After dilution in 1 ml 0.5 N HBr and
subsequent evaporation, it was diluted again in 0.5 N HBr and centrifuged. Lead was
separated on anion exchange columns containing 100 μl resin (Biorad AG1-X8, 200-400
mesh). The samples were dissolved in 0.3 ml 2 N HCl and washed into the columns,
afterward rinsed with 1ml 0.5N HBr and finally eluted with 1 ml 6N HCl. To get rid of alkalis
and earth-alkalis, the separation process was repeated. The entire processing was carried out
within a laminar flow box to exclude contamination by lead which is bound to dust particles.
Pb was mounted on Re double filament using silica-gel. Lead isotopes were corrected to NBS
981. Normalization of our data to recommended values was performed using a mass
fractionation factor of 0.122%. From measurements of twenty one standards gave a mean of
206
Pb/204Pb= 16.90 ± 0.01, 207Pb/204Pb= 15.44 ± 0.02, 208Pb/204Pb= 37.53 ± 0.05, total error
(2σ) < 0.1% were determined. Total blank were 0.29 ng. Ionisation temperature for Pb
measurement was held constant at 1200°C.
108
6 References
Acosta, J., 2004, Facies y ambientes sedimentarios de la formacion Moquegua superior en los
alrededores de la ciudad de Tacna: In: J. Jacay & T. Sempere (eds.), Nuevas contribuciones del IRD y
sus contrapartes al conocimiento geológico del sur del Perú, Sociedad Geológica del Perú, Publicación
Especial n° 5, p. 183-186.
Aitcheson, S.J., Forrest, A.H., 1994, Quantification of crustal contamination in open magmatic
systems: Journal of Petrology, v. 35, p. 461-488.
Aitcheson, S.J., Harmon, R.S., Moorbath, S., Schneider, A., Soler, P., Soria, E.E., Steele, G.,
Swainbank, I., Wörner, G., 1995, Pb isotopes define basement domains of the Altiplano, Central
Andes: Geology, v. 23, p. 555-558.
Allmendinger, R., Gonzalez, G., Yu, J., Hoke, G., Isacks, B., 2005, Trench-parallel shortening in the
northern Chilean forearc; tectonic and climatic implications: Geological Society of America Bulletin,
v. 117, p. 89-104.
Allmendinger, R.W., Gubbels, T., 1996, Pure and simple shear plateau uplift, Altiplano-Puna,
Argentina and Bolivia: Tectonophysics, v. 259, p. 1-13.
Allmendinger, R.W., Jordan, T.E., Kay, S.M. and Isacks, B., 1997, The evolution of the Altiplano-
Puna plateau of the Central Andes: Annual Rewiews of Earth and Planetary Sciences, v. 25, p. 139-
174.
Angermann, D., Klotz, J., Reiber, C., 1999, Space-geodetic estimation of the Nazca-South American
Euler vector: Earth and Planetary Science Letters, v. 171, p. 329 - 334.
Anthes, G., 1993, Haupt-, Spurenelement und Isotopengeochemie an den Oberkretazischen und
Tertiaren Granitoiden der Präkordillere Nordchiles zwischen 10° und 22°S, MS Thesis (Unpublished)
Universität Mainz.
Astini, R.A., Benedetto, J.L.,Vaccari, N.E., 1995, The early Paleozoic evolution of the Argentine
Precordillera as a Laurentian rifted, drifted and collided terrane: a geodynamic model: Geol. Soc. Am.
Bull., v. 107, p. 253-273.
Ayuso, R.A., Bevier, M.L., 1991, Regional differences in Pb isotopic compositions of feldspars in
plutonic rocks of the northern Appalachian Mountains USA, and Canada: A geochemical method of
terrain correlation: Tectonics, v. 10, p. 191-212.
Babeyko, A.Y., Sobolev, S.V., Trumbull, R.B., Oncken, O., Lavier, L.L., 2002, Numerical models of
crustal scale convection and partial melting beneath the Altiplano-Puna plateau: Earth Planet. Sci.
Lett., v. 199, p. 373- 388.
Baby, P., Rochat, P., Mascle, G., Herail, G., 1997, Neogene shortening contribution to crustal
thickening in the back arc of the Central Andes: Geology, v. 25, p. 883-886.
Bahlburg, H., Hervé, F., 1997, Geodynamic evolution and tectonostratigraphic terranes of
northwestern Argentina and Northern Chile: Geological Society of America Bulletin, v. 109, p. 869-
884.
Barreiro, B.A., 1984, Lead isotope and Andean magmagenesis in Harmon, R.S., and Barreiro, B.A.,
eds., Andean magmatism, chemical and isotopic constraints: Cheshire, UK, Shiva, p. 21-30.
109
Barreiro, B.A., Clark, A.H., 1984, Lead isotopic evidence for evolutionary changes in magma crust
interaction, Central Andes, southern Peru: Earth Planet. Sci. Lett., v. 69, p. 30-42.
Beate, B., Monzier, M., Springs, R., Cotton, J., Bourdon, E., Eissen, J. P., 2001, Mio-Pliocene adakite
generation related to flat subduction in southern Ecuador: Quinsacocha Volcanic center: Earth Planet.
Sci. Lett., v. 192, p. 561-570.
Beck, S.L., Zandt, G., Myers, S.C., Stephen, C., Wallace, T.C., Silver, P.G. , 1996, Crustal thickness
variations in the central Andes. Geology, v. 24, p. 407-411.
Beck, S.L., Zandt, G. , 2002, The nature of orogenic crust in the Central Andes: J. Geophys. Res., v.
107 doi:10.1029/2000JB000124.
Bellon, H., Lefévre, C., 1976, Données géochronométriques sur le volcanisme andin dans le sud du
Pérou, v. Implications volcano-tectoniques C.R. Acad. Sci. Paris 238, Ser. D, 1-4.
Bock, B., Bahlburg, H., Wörner, G., Zimmermann, U., 2000, Tracing crustal evolution in the southern
Central Andes from Late Precambrian to Permian using Nd and Pb isotopes: Journal of Geology, v.
108, p. 515- 535.
Bogdanic, T., Espinoza, S., 1994, Tectono-sedimentary evolution of the Cretaceous-Early Tertiary and
metallogenic scheme of Northern Chile: In: Salfity, J. (ed.) Cretaceous Tectonics of the Andes.
Braunschweig: Vieweg, p. 213-265.
Boily, M., Brooks, C., Ludden, J.N., 1989, Chemical and Isotopic Evolution of the Coastal Batholith
of Southern Peru: Journal of Geophysical Research, v. 94, p. 483-498.
Boric, R., Díaz, F., Maksaev, V., 1990, Geología y yacimientos metalíferos de la Región de
Antofagasta: ServicioNacional de Geología y Minería, Boletín, p. 246, 2 mapas escala 1:500.000.
Bosch, M., Rodriguez, I., 1992, North Venezuelan collisional crustal block; the boundary between the
Caribbean and South American plates: Journal of South American Earth Sciences, v. 6, p. 133-143.
Bourdon, B., Wörner, G., Zindler, A., 2000, U-series evidence for crustal involvement and magma
residence times in the petrogenesis of Parinacota Volcano, Chile: Contributions to Mineralogy and
Petrology, v. 139, p. 458-469.
Bourdon, E., Eissen, J.P., Monzier, M., Robin, C., Martin, H., Hall, M.L., 2002, Adakite-like lavas
from Antisana Volcano (Ecuador): evidence for slab melt metasomatism beneath the Northern Andean
Zone: J. Petrol, v. 43, p. 199-217.
Cahill, T., Isacks, B.L., 1992, Seismicity and shape of the subducted Nazca plate: Journal of
Geophysical Research, v. 97, p. 17, 503-17, 529.
Carlier, G., Lorand, J.P., Liegeois, J.P., Fornari, M., Soler, P., Carlotto, V., Cardenas, J., 2005,
Potassic-ultrapotassic mafic rocks delineate two lithospheric mantle blocks beneath the southern
Peruvian Altiplano: Geology, v. 33, p. 601-604.
Carlotto, V., 1998, Evolution andine et raccourcissement au niveau de Cusco (13°-16°S, Pérou),
Thése de doctorat, Université de Grenoble, France, p. 159.
Cassard, D., Chauvet, A., Bailly, L., Llosa, F., Rosas, J., Marcoux, E., Lerouge, C., 2000, Structural
control and K/Ar dating of the Au-Ag epithermal veins in the Shila Cordillera, southern Peru: C. R.
Acad. Sci. Paris, v. 330, p. 23-30.
110
Charrier, R., Reutter, K.J., 1994, The Purilactis Group of Northern Chile: Boundary Between Arc and
Backarc from Late Cretaceous to Eocene In: Reutter, K.J., Scheuber, E. & Wigger, P. (eds.): Tectonics
of the Southern Central Andes: Springer Verlag, Berlin, p. 189-202.
Chiaradia, M., Fontboté, L., 2002, Lead isotope systematics of Late Cretaceous – Tertiary Andean arc
magmas and associated ores between 8°N and 40°S: evidence for latitudinal mantle heterogeneity
beneath the Andes: Terra Nova, v. 14, p. 337–342.
Clark, A.H., Farrar, E., Kontak, D.J., Langridge, R.J., Arenas, M.J., France, L.J., McBride, S.L.,
Woodman, P.L., Wasteneys, H.A., Sandeman, H.A., Archibald, D.A., 1990a, Geologic and
geochronologic constraints on the metallogenic evolution of the Andes of southeastern Peru:
Economic Geology, v. 85, p. 1520-1583.
Clark, A.H., Tosdal, R.M., Farrar, E., Plazolles, A., 1990b, Geomorphologic environment and age of
supergene enrichment of the Cuajone, Quellaveco, Toquepala Porphyry Copper deposits, southeastern
Peru: Economic Geology, v. 85, p. 1604-1628.
Cobbing, E.J., Ozard, J.M., Snelling, N.J., 1977, Reconnaissance geochronology of the crystalline
basement rocks of the Coastal Cordillera of southern Peru: Geol. Soc. Am. Bull., v. 88, p. 241-246.
Coira, B., Davidson, J., Mpodozis, C., Ramos, V., 1982, Tectonic and magmatic evolution of the
Andes of northern Argentina and Chile: Earth Sci. Rev, v. 18, p. 303-332.
Coira, B., Kay, S.M., Viramonte, J., 1993, Upper Cenozoic magmatic evolution of the Argentine
Puna_ A model for changing subduction geometry: International Geology Review, v. 35, p. 677-720.
Dalmayrac, B., Laubacher, G., Marocco, R., 1977, Géologie des Andes Peruviennes, Caracteres
Generaux de L'évoution géologique des Andes Peruviennes, Thése Universite des Sciences et
Techniques du Languedoc. p. 361.
Dalziel, I.W.D., Forsythe, R.D., 1985, Andean evolution and the terrane concept: In
Tectonostratigraphic Terranes of the Circum-Pacific Region; Circum-Pacific council for Energy and
Mineral Resources, Earth Sci. Series, v. 1, p. 565-581.
Damm, K.W., Pichowiak, S., Harmon, R.S., Todt, W., Omaniri, R., Niemeyer, H., 1990, Pre-Mesozoic
evolution of the Central Andes-the basement revisited: Geol. Soc. Am. Spec., v. 241, p. 101-126.
Damm, K.W., Harmon, R., Kelley, S., 1994, Some isotopic and geochemical constraints on the origen
and evolution of the Central Andes basement (19°-24°): In Reutter, KJ, Scheuber E & Wigger PJ (eds)
Tectonic of the Southern Central Andes. Springer Verlag, Berlin, p. 263-276.
Dasch, J., 1981, Lead isotopic composition of metalliferous sediments from Nazca plate: Geological
Society of America Memoir, v. 154, p. 199-210.
Davidson, J.P., McMillan, N. J., Moorbath, S., Wörner, G., Harmon, R.S., Lopez-Escobar, L., 1990,
The Nevados de Payachata Volcanic Region (18°S/69°W, N. Chile) II: Evidence for Widespread
Crustal Involvement in Andean Magmatism: Contributions to Mineralogy and Petrology, v. 105, p.
412-432.
Davidson, J.P., Harmon, R.S., Wörner, G., 1991, The source of the Central Andes magmas: some
considerations: Geol. Soc. Am. Andean magmatism and its tectonic setting, v. 265, p. 233-244.
Davidson, J.P., de Silva, S., 1992, Volcanic rocks from the Bolivian Altiplano: Insights into crustal
structure, contamination, and magma genesis in the central Andes: Geology, v. 20, p. 1127-1130.
111
De Silva, S.L., 1989, Altiplano-Puna volcanic complex of the central Andes: Geology, v. 17, p. 1102-
1106.
De Silva, S.L., Francis, P.W., 1990, Potentially active volcanoes of Peru - Observations using Landsat
Thematic Mapper and Space Shuttle imagery: Bulletin of Volcanology, v. 52, p. 286-301.
—, 1991, Volcanoes of the Central Andes: Heidelberg, Springer-Verlag, 232 p.
Defant, M.J., Drummond, M.S., 1990, Derivation of some modern arc magmas by melting of young
subducted lithosphere: Nature, v. 347.
DePaolo, D.J., Wasserburg, G.J., 1976, Nd isotope variations and petrogenetic models: Geophys. Res.
Lett., v. 3, p. 249-252.
DePaolo, D.J., 1981, Trace element and isotopic effects of combined wallrock assimilation and
fractional crystallization: Earth and Planetary Science Letters, v. 53, p. 189-202.
DePaolo, D.J., Linn, A.M., Schubert, G., 1991, The Continental crustal age distribution: Methods of
determining mantle separation ages from Sm-Nd isotopic data and application to the southwestern
Unitet States: Journal of Geophysical Research, v. 96, p. 2071-2088.
Dorbath, C., Granet, M., 1996, Local earthquake tomography of the Altiplano and Eastern Cordillera
of northern Bolivia: Tectonophysics, v. 259, p. 117-136.
Dostal, J., Zentilli, M., Caelles, J.C., Clark, A.H., 1977, Geochemistry and origen of volcanic rocks of
the Andes (26-28°S): Contrib. Mineral. Petrol., v. 63, p. 113-128.
Entenmann, J., 1994, Magmatic evolution of the Nevados de Payachata Complex and the petrogenesis
of basaltic andesites in the Central Volcanic Zone of northern Chile, PhD Thesis, Universität Mainz, p.
184.
Eusterhues, K., 1993, Geochemie miokaener Vulkanite Nordchiles (18°-20°S), MS Thesis

(Unpublished) Universität Mainz.
Flores, A., Sempere, T., Fornari, M., 2004, Síntesis actualizada de la estratigrafía del Cenozoico en el
extremo sur del Peru: XII Congreso Peruano de Geología. Resúmenes Extendidos. Sociadad
Geológica del Perú.
Flores, A., Acosta, J., Bedoya, C., Sempere, T., 2005, Oligocene-Neogene tectonics and sedimentation
in the forearc of southern Peru, Tacna area (17.5º-18.5ºS): in VI International Symposium on Andean
Geodynamics, September, Barcelona, Spain, p. 269-262.
Fornari M., B.E., Espinoza F., Ibarra I., Jiménez N., Mamani M., 2002, Ar-Ar Dating of Late
Oligocene-Early Miocene Volcanism in the Altiplano: V International Symposium on Andean
Geodinamics, Toulouse-French, p. 223-226.
Francis, P.W., Hawkesworth, C.J., 1994, Late Cenozoic rates of magmatic activity in the Central
Andes and their relationships to continental crust formation and thickening: J. Geol. Soc. (London), v.
151, p. 845-854.
Garbe-Schönberg, C.D., 1993, Simultaneous determination of 37 trace elements in 28 international

rock standards by ICP-MS: Geostandards Newsletter, v. 17, p. 81-93.
Garrison, J.M., Davidson, J.P., 2003, Dubious case for slab melting in the Northern volcanic zone of
the Andes: Geology, v. 31, p. 565 - 568.
112
Giese, P., Scheuber, E., Schilling, F., Schmitz, M., Wigger, P., 1999, Crustal thikening processes in
the Central Andes and the different natures of the Moho-discontinuity: J. South Am. Earth Sci., v. 12,
p. 201-220.
Gil, W., 2001, Evolution latérale de la déformation d'un front orogénique: Example des Bassins
subandins entre 0° et 16°S: Toulouse, Phd Thesis, Université Paul Sabatier. p.150.
Gill, J., 1981, Orogenic andesites and plate tectonics: Springer Verlag, Berlin, p. 330.
Gorring, M.L., Kay, S.M., Zeitler, P.K., Ramos, V.A., Rubiolo, D., Fernández, M.I. and Panza, J.L.,
1997, Neogene Patagonian plateau lavas: continental magmas associated with ridge collision at the
Chile Triple Junction: Tectonics, v. 16, p. 1-17.
Gregory-Wodzicki, K.M., 2000, Uplift history of the central and northern Andes: A review: Geol. Soc.
Am. Bull., v. 112, p. 1091-1105.
Guerra, E., 1993, Petrographie und Geochemie der magmatischen Gesteine und Petrographie der
Sedimentgesteine des Murmutani Escarpment in the Präkordillere Nordchiles: Diploma Tesis
(Unpublished), Universität Maiz, p. 115.
Gunnesch, K.A., Baumann, A., Gunnesch, M., 1990, Lead isotope variations across the central
Peruvian Andes: Economic Geology, v. 85, p. 1384-1401.
Gutscher, M.A., Olivet, J.L., Aslanian, D., Eissen, J.P., Maury, R., 1999, The ‘‘lost Inca Plateau’’:
Cause of flat subduction beneath Peru?: Earth Planet. Sci. Lett., v. 171, p. 335-341.
Gutscher, M.A., Maury, R., Eissen, J.P., and Bourdon, E., 2000, Can slab melting be caused by flat
subduction?: Geology, v. 28, p. 535-538.
Gutscher, M.A., 2002, Andean subducting styles and their effect on thermal structure and interplate
coupling: J. South Am. Earth Scien., v. 15, p. 3-10.
Hampel, A., 2002, The migration history of the Nazca Ridge along the Peruvian active margin; a re-
evaluation: Earth and Planetary Science Letters, v. 203, p. 665-679.
Harmon, R.S., Hoefs, J., 1984, Oxygen isotope ratios in Late Cenozoic Andean volcanics: in Harmon,
R.S. & Barreirro, B.A. (eds.), Andean magmatism, chemical and isotopic constraints. Shiva Publishing
Ltd, p. 9-21.
Harmon, R.S., Barreiro, B.A., Moorbath, S., Hoefs, J., Francis, P.W., Thorpe R.S., Déruelle, B.,
McHugh, J., Viglino, J.A. , 1984, Regional O-, Sr- and Pb relationships in Late Cenozoic calc-alkaline
lavas of the Andean Cordillera: J. Geol. Soc. London, v. 141, p. 803-822.
Haschke, M., Günther, A., 2003, Balancing crustal thickening in arcs by tectonic vs. magmatic means:
Geology, v. 31, p. 933-936.
Haschke, M.R., Siebel, W., Günther, A., and Scheuber, E., 2002, Repeated crustal thickening and
recycling during the Andean orogeny in north Chile (21°-26°S): Journal of Geophysical Research, v.
107.
Heinrichs, H., Herrmann, A.G., 1990, Praktikum der Analytischen Geochemie: Springer, Berlin.
Heumann, A., 1993, Geochemistry and Petrology of Upper Cretaceous and Tertiary Granitoids in
northern Chile 18°-22°S, MS Thesis (Unpublished), Universität Mainz.
113
Hildreth, W., Moorbath, S., 1988, Crustal contributions to arc magmatism in the Andes of central
Chile: Contributions to Mineralogy and Petrology, v. 98, p. 455-489.
Hindle, D., Kley, J., Oncken, O., Sobolev, S., 2005, Crustal balance and crustal flux from shortening
estimates in the central Andes: Earth Planet. Sci. Lett., v. 230, p. 113-124.
Horn, S., 1992, Intra- und intervulkanische Variationen entlang der Zentralen Vulkanzone in Nordchle
17-22°S. Petrographische und geochemische Untersuchungen: MS Thesis (Unpublished), Universität
Mainz, p. 145.
Husson, L., Sempere, T., 2003, Thickening the Altiplano crust by gravity-driven crustal channel flow:
Geophys. Res. Lett., v. 30 (5), 1243, doi: 10.1029/2002GL016877.
Isacks, B.L., 1988, Uplift of the Central Andean plateau and bending of the Bolivian Orocline: J.
Geophys. Res., v. 93, p. 3211-3231.
James, D.E., 1982, A combined O, Sr, Nd and Pb isotopic and trace element study of crustal
contamination in central Andean lavas, I. Local geochemical variations: Earth Planet. Sci. Lett., v. 57,
p. 47-62.
James, D.E., Sacks, I. S., 1999, Cenozoic formation of the central Andes: a geophysical perspective in
Geology and Ore Deposits of the Central Andes: (SpecialPublication no. 7), B. J. Skinner, ed.,
Littleton, Society of Economic Geologists, p. 1-25.
Jordan, T.E., Isacks, B.L., Allmendinger, R.W., Brewer, J.A., Ramos, V.A., Ando, C.J. , 1983,
Andean tectonic related to the geometry of the subducted Nazca Plate: Geol. Soc. Am. Bull., v. 94, p.
341-361.
Jordan, T.H., 1988, Structure and formation of the continental tectosphere: J. Petrology, special
lithosphere issue, Menzies, M.A. (Ed.), p. 11-37.
Kamenov, G., Macfarlane, A.W., Riciputi, L.R., 2002, Sources of lead in the San Cristobal, Pulacayo,
and Potosi mining districts, Bolivia, and a reevaluation of regional ore lead isotope provinces:
Economic Geology, v. 97, p. 573-592.
Kaneoka, I., Guevara, C., 1984, K-Ar age determinations of late Tertiary and Quaternary Andean
volcanic rocks, southern Peru: Geochemical Journal, v. 18, p. 233-239.
Kay, R.W., 1977, Geochemical constraints on the origen of Aleutian magmas: In Talwani, M. Pittman
III W.C. (eds) Island Arcs, Back-arc Basins, and Deep-sea Trenches. Amer Geophys Union, Maurice
Ewing Series v. 1, p. 229-242.
Kay, S.M., Coira, B., Viramonte, J., 1994, Young mafic back arc volcanic rocks as indicators of
continental lithospheric delamination beneath the Argentine Puna Plateau, Central Andes: American
Geophysical Union, v. 99, p. 24,323-24,339.
Kay, S.M., Mpodozis, C., Coira, B., 1999, Neogene magmatism, tectonism, and mineral deposits of
the Central Andes (22˚S to 33˚S): in Skinner, B. et al., eds, Geology and Mineral Deposits of Central
Andes, Society of Economic Geology Special Publication, v. 2, p. 1-35.
Kay, S.M., Mpodozis, C., Coira B., 1999, Neogene magmatism, tectonism and mineral deposit of the
Central Andes (22° to 33°S Latitude): In Skinner B.J. (ed) Geology and Ore Deposit of the Central
Andes. Soc. Econ. Geol. Spec. Pub. 7, p. 27-59.
114
Kay, S.M., 2002, Andean Adakite from slab melting, crustal thickening, and forearc subduction
erosion: in V International Symposium on Andean Geodynamics, September, Tolouse, France, p. 405-
408.
Kay, S.M., Godoy, E., Kurtz, A., 2005, Episodic arc migration, crustal thickening, subduction erosion,
and magmatism in the south-central Andes: Geol. Society of America, v. 117, p. 67-88.
Kay, S.M., Mpodozis, C., Ramos, V.A., and Munizaga, F., 1991, Magma source variations for mid-
late Tertiary magmatic rocks associated with a shallowing subduction zone and a thickening crust in
the central Andes (28 to 33°S): in Harmon, R.S., and Rapela, C.W., eds., Andean magmatism and its
tectonic setting: Geological Society of America Special Paper, v. 265, p. 113-137.
Kennan, L., Lamb, S., Rundle, C., 1995, K-Ar dates from the Altiplano and Cordillera Oriental of
Bolivia: Implications for Cenozoic stratigraphy and tectonic: Journal of South American Earth
Sciences, v. 8, p. 163-186.
Kistler, R.W., 1990, Two different lithosphere types in the Sierra Nevada: Geologic Society of
America Memoir, v. 174, p. 271-281.
Kley, J., Monaldi, C.R., 1998, Tectonic shortening and crustal thickness in the Central Andes: How
good is the correlation?: Geology, v. 26, p. 723 - 726.
Klinck, B.A., Ellison, R.A., Hawkins, M.P., 1986, The geology of the Cordillera Occidental and
Altiplano, West of the Lake Titicaca, southern Peru. Instituto de Geología Minería y Metalurgía,
Preliminary report: Lima, p. 353.
Kohlbach, I., 1999, Spatial and temporal variations in magma geochemistry along a W-E traverse at
18°-19°S, North Chile, MS Thesis (Unpublished), Universität Göttingen.
Kontak, D.J., Cumming, G.L., Krstic, D., Clark, A.H., Farrar, E., 1990, Isotopic composition of lead in
ore deposits of the Cordillera Oriental, southeastern Peru: Economic Geology, v. 85, p. 1584-1603.
Kyser, T., 1986, Stable isotopic variations in the mantle: In Valley, J.W., Taylor, H.P.Jr., O´Neil, J.R.
(eds.) Stable isotopes in high temperature geological processes. Rev. Mineral., v. 167, p. 41-164.
Lamb, S., Hoke, L., Kennan, L., Dewey, J., 1997, Cenozoic evolution of the Central Andes in the
Bolivia and Northern Chile. In: Burg, J.p., Ford, M. (Eds), Orogeny through time. Geol. Soc. Spec.
Publ., v. 121, p. 3237-264.
Laubacher, G., Bonhomme, M., Fornari, M., Herail, G., Vivier, G., 1984, Le bassin d' Ananea-
Ancocala, témoin de l'évolution plio-qauternaire des Andes sud-orientales du Pérou: In Dexiéme
RAST, Bordeaux, p. 336.
Leeman, W.P., 1983, The influence of crustal structure of compositions of subduction- related
magmas. Arc Volcanism: J. Volc. Geotherm. Res., v. 18, p. 561-588.
Lehmann, B., 1978, A Precambrian core sample from the Altiplano/Bolivia: Geologische Rundschau,
v. 67, p. 270-278.
Loewy, S.L., Connelly, J.N., Dalziel, I.W.D., 2004, An Orphaned Basement Block: The Arequipa-
Antofalla Basement of the Central Andean margin of South America: Geol. Soc. Am. Bull., v. 116, p.
171-187.
115
López-Escobar, L., Frey, F., Vergara, M., 1977, Andesites and high-alumina basalts from central-
south Chile High Andes: Geochemical evidence bearing on their petrogenesis: Contrib. Mineral.
Petrol., v. 63, p. 199-228.
Lork, A., Bahlburg, H., 1993, Precise U-Pb ages of monazites from the Faja Eruptiva de la Puna
Oriental and the Cordillera Oriental, NW Argentina: XII Congreso Geológico Argentino y II Congreso
de Exploración de Hidrocarburos Actas, v. IV, p. 1-6.
Lucassen, F., Fowler, C.M.R., Franz, G., 1996, Formation of magmatic crust at the Andean continental
margin during early Mesozoic: a geological and thermal model of the North Chilean Coast Range:
Tectonophysics, v. 262, p. 263- 279.
Lucassen, F., Franz, G., 1996b, Magmatic arc metamorphism: petrology and temperature history of
metabasic rocks in the Coastal Cordillera of northern Chile/Region Antofagasta: J. Metamorph. Geol.
Rundsch, v. 14, p. 249-265.
Lucassen, F., Thirwall, MF., 1998, Sm-Nd formation ages and mineral ages in metabasites from the
Coastal Cordillera, northern Chile: Geol. Rundsch, v. 86, p. 767-774.
Lucassen, F., Lewerenz, S., Franz, G., Viramonte, J., Mezger, K., 1999b, Metamorphism, isotopic
ages and composition of lower crustal granulite xenoliths from the Cretaceous Salta Rift, Argentina:
Contributions to Mineralogy and Petrology, v. 134, p. 325- 341.
Lucassen, F., Becchio, R., Harmon, R., Kasemann, S., Franz, G., Trumbull, R., Wilke, H., Romer, R.,
Dulski, P., 2001, Composition and density model of the continental crust at an active continental
margin-the Central Andes between 21°S and 27°S: Tectonophysics, v. 341, p. 195-223.
Lucassen, F., Escayola, M., Romer, R.L., Viramonte, J., Koch, K., Franz G., 2002, Isotopic
composition of the Late Mesozoic basic and ultrabasic rocks from the Andes (23-32°S) - Implications
for the Andean mantle: Mineral Petrologie, v. 143, p. 336 - 349.
Macfarlane, A.W., Marcet, P., LeHuray, A.P., Petersen, U., 1990, Lead isotope provinces of the
central Andes inferred from ores and crustal rocks: Economic Geology, v. 85, p. 1857-1880.
Mamani, M., Ibarra, I., Carlier, G., & Fornary, M., 2004, Petrología y geoquímica del magmatism
alcalino de la zona noroeste del Altiplano peruano (departamento de Puno): In: J. Jacay & T. Sempere
(eds.), Nuevas contribuciones del IRD y sus contrapartes al conocimiento geológico del sur del Perú,
Sociedad Geológica del Perú, Publicación Especial n° 5, p. 157-174.
Marocco, R., Delfaud, J., Lavenu, A., 1985, Ambiente deposicional de una cuenca contanintal
intramontaña andina: el Grupo Moquegua (sur de Perú) primeros resultados: Bol. Soc. Geol. Perú, v.
75, p. 73-90.
Marrett, R., Alllmendiger, R., Alonso, R., Drake, R., 1994, Late Cenozoic tectonic evolution of the
Puna Plateau and adjacent foreland, northwestern Argentine Andes: Journal of South America Earth
Sciences, v. 7, p. 179-207.
McDonough, W.F., Sun, S.-s.. Ringwood, A.E., Jagoutz, E., Hofmann, A.W., 1992, Potassium,
rubidium and cesium in the Earth and Moon and the evolution of the mantle of the Earth: Geochim.
Cosmochim Acta, v. 56, p. 1001-1012.
McDonough, W.F., Sun, S.-S., 1995, The composition of the Earth: Chem. Geol, v. 120, p. 223-253.
McQuarrie, N., DeCelles, P., 2001, Geometry and structural evolution of the central Andean back
thrust belt, Bolivia: Tectonics, v. 20, p. 669-692.
116
McQuarrie, N., 2002, Initial plate geometry, shortening variations, and evolution of the Bolivian
orocline: Geology, v. 30, p. 867-870.
Miller, D.M., Goldstein, S.L., Langmuir, C.H., 1994, Cerium/lead and lead isotope ratios in arc
magmas and the enrichment of lead in the continents: Nature, v. 368, p. 514-520.
Mlynarczyk, M.S.J., Williams-Jones, A.E. , 2005, The role of collisional tectonics in the metallogeny
of the Central Andean tin belt: Earth and Planetary Science Letters, v. 240, p. 656-667.
Morris, J.D., Leeman, W.P., Tera, F., 1990, The subducted component in island arc lavas: constraints
from Be isotopes and B-Be systematic: Nature, v. 344, p. 31-36.
Mpodozis, C., Hervé, F., Davidson, J., Rivano, S., 1983, Los granitoides de Cerros de Lila,
manifestaciones de un episodio intrusivo y termal del Paleozoico inferior en los Andes del Norte de
Chile: Revista Geológica de Chile, v. 18, p. 3-14.
Mukasa, S.B., 1986, Common Pb isotopic compositions of the Lima, Arequipa and Toquepala
segments in the Coastal Batholith, Peru: implication for magma genesis: Geochim. Cosmochim. Acta,
v. 50, p. 771-782.
Mukasa, S.B., Vidal, C.E., Injoque-Espinoza, J., 1990, Pb isotopes bearing on the Metallogenesis of
sulfide ore deposits in central and southern Peru: Economic Geology, v. 85, p. 1438-1446.
Müller, J.P., Kley, J., Jacobshagen, V., 2002, Structure and Cenozoic kinematics of the Eastern
Cordillera, southern Bolivia (21°S): Tectonics, v. 21 (5)1037, doi:10.1029/2001TC001340.
Myers, S., Beck, S., Zandt, G., Wallace, T., 1998, Lithospheric-scale structure across the Bolivian
Andes from tomographic images of velocity and attenuation for P and S waves: J. Geophys. Res., v.
103, p. 21233- 21252.
Norambuena, E., Leffer-Griffin, L., Mao, A.; Dixon, T., Stein, S., Sacks, S.; Ocola, L., Ellis, M. ,
1998, Space geodetic observations of Nazca-South America convergence across the Central Andes:
Science v. 279, p. 358-362.
O'Hara, M.J., Mathews, R.E., 1981, Geochemical evolution in an advancing, periodically replenished,
periodically tapped, continuously fractionating magma chamber: Journal of Geological Society of
London, v. 138, p. 237-277.
Oversby, V.M., 1978, Lead isotopes in Archean plutonic rocks: Earth and Planetary Science Letters, v.
38, p. 237-248.
Pacci, D., Hervé, F., Munizaga, F., Kawashita, K., Cordani, U., 1980, Acerca de la edad Rb-Sr
Precambrica de rocas de la Formación Esquistos de Belén, Departamento de Parinacota, Chile: Revista
Geológica de Chile, v. 11, p. 43-50.
Pardo-Casas, F., Molnar, P., 1987, Relative motion of the Nazca (Farallon) and South American plates
since Late Cretaceous time: Tectonics, v. 6/3, p. 233-248.
Pearce , J.A., 1983, Role of the sub-continental lithosphere in magma genesis at active continental
margins: In C. Hawkesworth, Norry, MJ (Editor), Continental basalts and xenoliths. Shiva, Nantwitch,
p. 230-272.
117
Petford, N., Atherton, M.P., Halliday, A.N., 1996, Rapid magma production rates, underplating and
remelting in the Andes: isotopic evidence from northern-central Peru (9-11ºS): Journal of South
American Earth Sciences, v. 9, p. 69-78.
Pichowiak, S., Buchelt, M., Damm, K.W., 1990, Magmatic activity and tectonic setting of early stages
of the Andean cycle in northern Chile: Geol. Soc. Am. Spec. Pap, v. 241, p. 127-144.
Pichowiak, S., 1994, Early Cretaceous magmatism in the Coastal Cordillera and the central depression
of North Chile: In Scheuber, E., Wigger, P.J. (eds) Tectonic of the Southern Central Andes; structure
and evolution of an active continental margin. Springer Verlag, Berlin, p. 203-217.
Pilger, R.H.J., 1984, Cenozoic plate kinematic, subduction and magmatism: South American Andes:
Journal of the Geol. Soc. London, v. 141, p. 793-802.
Plank, T., Langmuir, C.H., 1988, An evolution of the global variation in the major element chemistry
of arc basalt: Earth Planet. Sci. Lett., v. 90, p. 349-370.
Potts, P.J., 1987, Inductively coupled plasma-mass spectrometry. In: A Handbook of Silicate Rock
Analysis: Blackie, Glasgow, p. 575-586.
Prinz, P., Wilke, H.G., Hillebrandt, A.V., 1994, Sediment accumulation and subsidence history in the
Mesozoic marginal basin of northern Chile: In: K.J. Reutter, E. Scheuber and P.J. Wigger (Editors),
Tectonics of the Southern Central Andes. Springer, Heidelberg, p. 219-232.
Puig, A., 1988, Gelogical and metallogenic significance of the isotopic composition of lead in galenas
of the Chilean Andes: Economic Geology, v. 83, p. 843-858.
Ramos, V.A., Jordan, T.E., Allmendinger, R.w., Mpodozis, C., Kay, S.M., Cortes, J.E. and Palma, M.,
1986, Paleozoic terranes of the Central Argentine-Chilean Andes: Tectonics, v. 5, p. 855-880.
Ramos, V.A., 1988, Late Proterozoic-Early Paleozoic of South America - a collisional history:
Episodes, v. 11, p. 168-174.
Ramos, V.A., Kay, S.M., 1992, The Southern Patagonian plateau basalts: retroarc testimony of a ridge
collision, Argentina: in Oliver, R.A., Vatin-Perignon, N. and Laubacher, G., eds, Andean
Geodynamics Symposium, Grenoble, Tectonophysics, v. 205, p. 261-282.
Ramos, V.A., 1999, Plate Tectonic Setting of the Andean Cordillera: Episodes, v. 22, p. 183-190.
Ramos, V.A., Aleman, A., 2000, Tectonic evolution of the Andes. In: Cordani, U.G., et al. (Eds),
Tectonic Evolution of South America. 31st Int. Geol. Cong., Rio de Janeiro, p. 635-685.
Richards, J.P., Boyce, A.J., Pringle, M.S. , 2001, Geologic evolution of the Escondida area, northern
Chile: A model for spatial and temporal localization of porphyry Cu mineralization: Economic
Geology, v. 96, p. 271-305.
Rickwood, P., 1989, Boundary lines within petrologic diagrams which use oxides of major and minor
elements: Lithos, v. 22, p. 247-263.
Rodriguez, B.D., 1998, Regional crustal structure beneath the Carlin trend, Nevada based on deep
electrical geophysical measurements, U.S. Geological Survey Open-File p. 15-19.
Rogers, G., Hawkesworth, C.J., 1989, A Geochemical Traverse across the north Chilean Andes:
Evidence for Crustal Generation from the Mantle Wedge: Earth and Planetary Science Letters, v. 91,
p. 271-285.
118
Roperch, P., Sempere, T., Macedo, O., Arriagada, C., Fornari, F., Tapia, C., García, M., Laj, C., 2006,
Counterclockwise rotation of late Eocene-Oligocene fore-arc deposits in southern Peru and its
significance for oroclinal bending in the central Andes: Tectonics, v. 25, TC3010,
doi:10.1029/2005TC001882, 2006.
Rousse, S., Gilder, S., Fornari, M., Sempere, T., 2005, Insight into the Neogene tectonic history of the
northern Bolivian Orocline from new paleomagnetic and geochronologic data: Tectonics, v. 24,
TC6007, doi:10.1029/2004TC001760, 2005.
Salfety, J.A., Marquillas, R.A., 1981, Las unidades estratigráficas cretácicas del Norte de la República
Argentina. In: Volkheimer, W., Musacchio, E. (Eds.): Cuencas sedimentarias del Jurásico y Cretácico
de America del Sur, Comité Sudamericano del Jurásico y Cretácico, v. 1, p. 303-317.
Scheuber, E., Bogdanic, T., Jensen, A., Reutter, K.J. , 1994, Tectonic development of the north
Chilean Andes in relation to plate convergence and magmatism since the Jurassic, in Tectonics of the
Southern Central Andes: edited by K.-J. Reutter, E. Scheuber, and P. Wigger, pp. 121- 139, Springer-
Verlag, New York.
Schmitz, M., 1994, A balanced model of the southern central Andes: Tectonic, v. 13, p. 484-492.
Schreiber, U., Schwab, K., 1991, Geochemistry of Quaternary shoshonitic lavas related to the Calama-
Olacapato-El Toro lineament (NW Argentina): J. South Am. Aarth Sci., v. 4, p. 73-86.
Schröder, W., Wörner, G., 1996, Widespread Cenozoic ignimbrites in N-Chile, W-Bolivia and S-Peru
17°-20°S/71°-68°W. Stratigraphy extention, correlation and origin: 3rd ISAG, St. Malo, Andean
Geodynamics, ORSTOM Editions, Collection Colloques et Seminaires, p. 645-648.
Schweller, W.J., Kulm, L.D., Price, R.A., 1981, Tectonics, structure and sedimentary framework of
the Peru-Chile trench: in Kulm, L.D., Dymond, J., Desch, E.J., and Hussong, D.M., eds., Nazca Plate.
Crustal formation and Andean convergence. Geological Society of America Memoir, v. 154, p. 323-
351.
Sclater, J.G., Jaupart, C., Galson, D., 1980, The heat flow through oceanic and continental crust and
the heat loss of the Earth: Reviews of Geophysics and Space Physics, v. 18, p. 269-311.
Sébrier, M., Lavenu, A., Fornari, M., Soulas, J.P., 1988, Tectonic and uplift in Central Andes (Peru,
Bolivia and northern Chile) from Eocene to present: Géodynamique, v. 3, p. 85-106.
Sébrier, M., Soler, P., 1991, Tectonics and magmatism in the Peruvian Andes from Late Oligocene
time to the present: in Harmon, R.S. and Rapela, C.W., eds, Andean magmatism and its tectonic
setting, Geological Society of America Special Paper, v. 265, p. 259-276.
Sempere, T., 1990, Cuadros Estratigráficos de Bolivia: propuestas nuevas: Revista Tecnica de
Yacimientos Petrolíferos Fiscales Bolivianos, v. 11, p. 215-227.
Sempere, T., Marshall, L., Rivano, S., Godoy, E., 1994, Late Oligocene-Early Miocene compressional
tectonosedimentary episode and associated Lan-Mammal Faunas in the Andes of Central Chile and
adjacent Argentina (32-37°S): Tectonophysics, v. 229, p. 251-264.
Sempere, T., Fornari, M., Acosta, J., Flores, A., Jacay, J., Peña, D., Roperch, P., Taipe, E., 2004,
Estratigrafía, geocronología, paleogeografía y paleotectónica de los depósitos de antearco del sur del
Perú: XII Congreso Peruano de Geología. Resúmenes Extendidos. Sociadad Geológica del Perú.
Seyfried, H., Wörner, G., Uhlig, D., Kohler, I., 1995, Eine kleine Landschaftsgeschichte der Anden in
Nordchile: Wechselwirkungen, Jahrbuch 1994 der Universität Stuttgart, p. 60-71.
119
Shackleton, R.M., Ries, A.C., Coward, M.P., and Cobbold, P.R., 1979, Structure, metamorphism and
geochronology of the Arequipa Massif of coastal Peru: Journal of the Geological Society, London, v.
136, p. 195-214.
Siebel, W., Schnurr, W.B.W., Hahne, K., Kraemer, B., Trumbull, R.B., van den Bogaard, P.,
Emmermann, R., 2001, Geochemistry and isotope systematics of small- to medium-volume Neogene –
Quaternary ignimbrites in the southern central Andes: evidence for derivation from andesitic magma
sources: Chemical Geology, v. 171, p. 213– 237.
Sobolev, S., Babeyko, A., 1994, Modelling of mineralogical composition, density and elastic wave
velocities in the unhydrous rocks: Surveys in Geophysics, v. 15, p. 515-544.
Soler, P., Bohomme, M., 1990, Relations of magmatic activity to plate dynamics in the Central Peru
from Late Cretaceous to present: In Plutonism from Antartica to Alaska, Kay and Rapela eds.
Geological Society of America Special Paper, v. 241, p. 173-192.
Soler, P., Carlier, G., Fornari, M., Herail, G., 1992, An Alternative model for the origen and tectonic
significance of the Neogene and Quaternary shoshonitic volcanism of the Andes (abstract): Eos Trans.
AGU 73 14 Spring Meeting Suppl. , p. 341-342.
Somoza, R., 1998, Updated Nazca (Farallon)-South America relative motions during the last 40 My:
implications for the mountain building in the central Andean region: Journal of South American Earth
Sciences, v. 11, p. 211-215.
Spera, F.J., Bohrson, W.A., 2001, Energy-constrained open-system magmatic processes; I, General
model and energy-constrained assimilation and fractional crystallization (EC-AFC) formulation:
Journal of Petrology, v. 42, p. 999-1018.
Stacey, J.S., and Kramers, J.D., 1975, Approximation of terrestrial lead isotope evolution by a two
stages model: Earth Planetary Science Letters, v. 26, p. 207-221.
Stern, C., Killian. R., 1996, Role of the subducted slab, mantle wedge and continental crust in the
generation of adakites from the Andean Austral Volcanic Zone: Contributions to Mineralogy and
Petrology, v. 123, p. 263-281.
Stern, C., 2004, Active Andean volcanism: its geologic and tectonic setting: Rev. geol. Chile, v. 31, p.
161-206.
Stern, C.R., 1989, Pliocene to present migration of the volcanic front, Andean southern volcanic zone:
Revista Geológica de Chile, v. 16, p. 145-162.
—, 1990, Comment on "A geochemical traverse across the North Chilean Andes: evidence for crust
generation from the mantle wedge" by G. Rogers and C.J. Hawkesworth: Earth Planet. Sci. Lett., v.
101, p. 129-133.
—, 1991, Role of subduction erosion in the generation of Andean Magmas: Geology, v. 19, p. 78-81.
Suarez, M.a.B., C.M., 1992, Triassic rift-related sedimentary basins in northern Chile (24°-29°S): J. S.
Am. Earth Sci., v. 6, p. 109-121.
Sweeney, J.F., 1982, Structure and development of the polar margin of North America , In Scrutum,
R.A., ed., Dynamics of passive margins: American Geophysicla Union and Geological Association of
America Geodynamics Series, v. 6, p. 17-29.
120
Swenson, J., Beck, S., Zandt, G., 2000, Crustal structure of the Altiplano from broadband regional
waveform modeling: implications for the composition of thick continental crust: Journal of
Geophysical Research, v. 105, p. 607-621.
Tassara, A., 2006, Factors controlling the crustal density structure underneath active continental
margins with implications for their evolution: Geochemistry, Geophysics, Geosystems, v. 7(1): doi:
10.1029/2005GC001040. issn: 1525-2027.
Tassara, A., Götze, H.J., Schmidt, S., Hackney, R., 2006, Three-dimensional density model of the
Nazca plate and the Andean continental margin: In press (August 23, 2006) by Journal of Geophysical
Research.
Taylor, S., Maclennan, S., 1985, The continental crust; its composition and evolution, Blackwell
Scientific Publications, Oxford, England, 312 p.
Tebbens, S.F., Cande, S.C., 1997, Southeast Pacific tectonic evolution from early Oligocene to
present: J. Geophys. Res., v. 102, p. 12061-12084.
Thorpe, R., Francis, P., 1979, Variations in Andean andesite compositions and their petrogenetic
significance: Tectonophysics, v. 57, p. 57-70.
Thorpe, R., Francis, P., Harmon, R., 1981, Andean andesites and crustal growth: Philos. Trans. Soc.
London, v. 301(A), p. 305-320.
Thouret, J., Rivera, M., Wörner, G., Gerbe, MC., Finizola, A., Fornari, M., Gonzales, K., 2005,
Ubinas: The evolution of the historically most active volcano in Southern Peru: Bull. Volcanol., v. 67,
p. 557–589.
Tilton, G.R., Barreiro, B.A., 1980, Origen of lead in Andean calc-alkaline lavas, southern Peru:
Science, v. 210, p. 1245-1247.
Tosdal, R.M., Farrar, E., Clark, A.H., 1981, K-Ar geochronology of the late Cenozoic volcanic rocks
of the Cordillera Occidental, southernmost Peru: Journal of Volcanology and Geothermal Research, v.
10, p. 157-173.
Tosdal, R.M., and 15 others., 1993, Summary of Pb isotope compositions in epithermal precious-metal
deposits, Orcopampa area of southern Peru, Berenguela area of western Bolivia, and Maricunga belt in
north-central Chile, in investigaciones de metales preciosos en el complejo volcanico Neogenico-
Cuaternario de los Andes Centrales: La paz, Servicio Geológico de Bolivia, Servicio Nacional de
Geología y Minería, Chile, Instituto Geológico Minero y Metalurgico, Perú, and U.S. Geological
Survey, p. 47-55.
Tosdal, R.M., Gibson, P.C., Noble, D.C., 1995, Metal sources for Miocene precious-metal veins of the
Orcopampa, Shila, Cailloma and Arcata mining districts, southern Peru: Sociedad Geológica del Perú,
Volumen jubilar Alberto Benavides. Lima, p. 311-326.
Tosdal, R.M., 1996, The Amazon-Laurentian connection as viewed from the Middle Proterozoic rocks
in the central Andes, western Bolivia and northern Chile: Tectonics, v. 15, p. 827-842.
Tosdal, R.M., Wooden, J.L., Bouse, R.M., 1999, Pb isotopes, Ore deposits, and Metallogenic
Terranes: In Application of radiogenic isotopes of ore deposits research and exploration. Lambert,
D.D., Ruiz, J. (eds). Reviews in Economic Geology, v. 12, p. 1-28.
Trumbull, R.B., Wittenbrink, R., Hahne, K., Emmermann, R., Büsch, W., Gerstenberger, H., Siebel,
W., 1999, Evidence for Late Miocene to Recent contamination of arc andesites by crustal melts in the
121
Chilean Andes (25°–26°S) and its geodynamic implications: Journal of South American Earth,
Sciences, v. 12, p. 135– 155.
Viramonte, J.G., Kay, S.M., Becchio, R., Escayola, M., Novitski, I., 1999, Cretaceous rift related
magmatism in central-western South America: J. S. Am. Earth. Sci., v. 12, p. 109-121.
Vivallo, W., Henriquez, F., 1998, Génesis común de los yacimientos estratoligados y vetiformes de
cobre del Jurásico Medio a Superior en la Cordillera de la Costa, Región de Antofagasta, Chile: Rev.
geol. Chile, v. 25, p. 199-228.
Wasteneys, H.A., Clark, A.H., Farrar, E., Langridge, R.J., 1995, Grenvillian granulite-facies
metamorphism in the Arequipa Massif, Peru: A Laurentia-Gondwana link: Earth and Planetary
Science Letters, v. 132, p. 63-73.
Watts, A.B., Lamb, S.H., Fairhead, J.D., Dewey, J.F., 1995, Lithospheric flexure and bending of the
Central Andes: Earth Planet. Sci. Lett., v. 134, p. 9-21.
Whitman, D., Isacks, B., Kay, S., 1996, Lithospheric structure and along-strike segmentation of the
Central Andean Plateau: seismic Q, magmatism, flexure, topography and tectonics: Tectonophysics, v.
259, p. 29-40.
Wilson, M., 1989, Igneous Petrogenesis: A Global Tectonic Approach, Chapman & Hall, 466 p.
Wooden, J.L., Stacey, J.S., Howard, K.A., Doe, B.R., Miller, D.M., 1988, Lead isotope evidence for
the formation of Proterozoic crust in the southwestern Unitet States, in Ernst,W.G., ed.:
Metamorphism and crustal evolution, western conterminous Unitet Statet; Rubey, Englewood Cliffs,
new Jersey Prentice-Hall, v. 7, p. 68-86.
Wooden, J.L., DeWitt, E., 1991, Pb siotope evidence for the boundary between the Early Proterozoic
Mojave and the Central Arizona crustal provinces in western Arizona: Arizona Geological Society
Digest, v. 19, p. 27-50.
Wooden, J.L., Tosdal, R.M., Kistler, R.W., 1998, Pb-isotope mapping of crustal structure in the
northern Great Basin and relationship to Au deposit trends: U.S. Geological Survey Open-File Report
98-338, p. 20-33.
Wörner, G., Harmon, R.S., Davison, J.D., Moorbath, S., Turner, T.L., McMillan, N., Nye, C., López-
Escobar, L., Moreno, H., 1988, The Nevados de Payachata Volcanic Region 18°S/69°W, N. Chile. I.
Geological, geochemical and isotopic observations: Bulletin of volcanology, v. 30, p. 287-303.
Wörner, G., Moorbath, S., Harmon, RS., 1992, Andean cenozoic volcanics reflect basement isotopic
domains: Geology, v. 20, p. 1103-1106.
Wörner, G., López-Escobar, L., Moorbath, S., Horn, S., Entenmann, J., Harmon, R.S., Davison, J.D.,
1992, Variaciones Gequímicas, locales y regionales, en el arco volcánico Andino del Norte de Chile
(17°30´-22°00´S): revista Geológica de Chile, v. 19, p. 37-56.
Wörner, G., Moorbath, S., Entenmann, J., Harmon, R.S., Davison, J.D., López-Escobar, L., 1994,
Large geochemical variation along the central Arc of northern Chile (17.5-22°S): In Reutter, K.J.,
Scheuber, E., Wigger, P.J., (editors). Tectonics of the Southern Central Andes. Structure and evolution
of an active continental margin. Springer, Berlin, p. 77-92.
Wörner, G., Hammerschmidt, K., Hemjes-Kunst, F., Lezaun, J., Wilke, H., 2000, Geochronology (40
Ar-39Ar-, K-Ar-, and He-exposure-) ages of Cenozoic magmatic rocks from Northern Chile (18°-
122
22°S). Implications for magmatism and tectonic evolution of the central Andes: Rev. Geol. Chile, v.
27, p. 205-240.
Wörner, G., Lezaun, J., Beck, A., Heber, V., Lucassen, F., Zinngrebe, E., Rössling, R., and Wilke, H.
G., 2000, Precambrian and Early Paleozoic evolution of the Andean basement at Belén (northern
Chile) and Cerro Uyarani (western Bolivia Altiplano): J. S. Am. Earth Sci. Rev., v. 13, p. 717-737.
Yañez, G., Cembrano, J., Pardo, M., Ranero, C., Selles, D., 2002, The Challenger-Juan Fernandez-
Maipo major tectonic transition of the Nazca-Andean subduction system at 33 - 34°S: Geodynamic
evidence and implications: J. S. Am. Earth Sci., v. 15, p. 23-38.
Yuan, X., S. V. Sobolev, R., Kind, O., Oncken, and Andes Seismology Group, 2000, New constraints
on subduction and collision processes in the central Andes from comprehensive observations of P to S
converted seismic phases: Nature, v. 408, p. 958-961.
Yuan, X.S., S. V.; Kind, R., 2002, Moho topography in the central Andes and its geodynamic
implication: Earth and Planetary Science Letters, v. 199, p. 389-402.
Zandt, G., Beck, S.L., Ruppert, S.R., Ammon, C.J., Rock, D., Minaya, E., Wallace, T.C., Silver, P.G.,
Anomalous crust of the Bolivian Altiplano, Central Andes; constraints from broadband regional
seismic waveforms: Geophysical Research Letters, v. 23, p. 1159-1162.
Zartman, R.E., Doe, B.R., 1981, Plumbotectonics - The model: Tectonophysics, v. 75, p. 135-162.
Zartman, R.E., Haines, S.M., 1988, The plumbotectonic model for Pb isotope systematics among
major terrestrial reservoirs-a case for bi-directional transport: Geochimica et Cosmochimica acta, v.
52, p. 1327-1339.
Zbar, B., 1991, Die Genese Mafischer Einschluesse in den Andesiten der Zentralen Vulkanzone
(Nordchile), Diploma thesis, Universtiät Bochum.
123
Lebenslauf
Name Mamani-Huisa
Vorname Mirian-Irene
Geboren am 03-05-1975
in Cuyocuyo, Peru
Schulausbildung
1983-1986 Grundschule Santa Catalina de Siena, Juliaca-Peru.
1986-1992 Gymnasium Javier Heraud, Laberinto-Peru
Universität Ausbildung
1993-1998 Bachelor in Geologie
San Antonio Abad del Cusco
Fakultät für Geowissenschaften und Geographie.
Cusco, Peru
1998-2000 Diplom in Ing. Geologie

San Antonio Abad del Cusco
Cusco, Peru
2003 Beginn der Promotion an der Georg-August-Universität Göttingen

Göttingen, Deutschland
2006 Abschluss der vorliegenden Dissertation

Variation MagmaComposit CentralAndes MMamani

Uploaded by

Copyright:

Available Formats

Variation MagmaComposit CentralAndes MMamani

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Variation MagmaComposit CentralAndes MMamani

Uploaded by

Copyright:

Available Formats

Variations in magma composition in time and

space along the Central Andes (13°S-28°S)

zur Erlangung des Doktorgrades

Referent: Prof. Dr. G. Wörner

3.3 Geochemistry of Meso-Cenozoic magmatic rocks in the CA ........................................ 60

5 Appendix ............................................................................................................................. 104

6 References ........................................................................................................................... 108

Chapter 3: Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the

Chapter 3: Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the

Esta disertación investiga la diversidad de la química en espacio y tiempo de los magmas en

Chapter 3: Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the

This dissertation studies the geochemistry and isotope composition of Meso-Cenozoic

Chapter 1 is an introduction to the dissertation discussing the open questions related

1.2 Working Hypothesis

1.3 Previous work

1.3.1 The Andes

a) Northern Volcanic Zone (NVZ) extending from 2°S to 5°N

1.3.2 The Central Volcanic Zone (CVZ)

1.3.2 Pre-Andean history

Outcrops Rock type Protolith age (Ma) Metamorphic age (Ma)

1.3.3 The Andean Cycle

1.3.4 Uplift of the Central Andes and subducting Nazca plate

Nazca Plate subduction

1.3.5 Andean volcanism

1.3.6 Magma genesis

Enriched sub-continental lithospheric mantle

of erupted lavas are established. Furthermore, ascending magmas can subsequently be

2 Geochemical variations in south Peruvian volcanic rocks (13°S-18°S): The role of

of sequence-stratigraphical studies as well as mining and exploration in volcanics of the

2.2 Tectonic Setting

2.3 Description of Cenozoic units and sampled volcanic centers

2.3.1 Middle Eocene - Lower Oligocene, ~ 45 to ~30 Ma (Lower Moquegua Formation,

2.3.3 Upper Miocene – Lower Pliocene, ~10 to ~3 Ma (Lower Barroso Formation)

2.3.4 Pliocene, ~3 to ~2 Ma (Sencca Formation)

2.3.5 Upper Pliocene to Pleistocene, 3 to 0.8 Ma (Upper Barroso Formation)

2.3.6 Holocene Volcanoes < 0.01 Ma

Other Pleistocene to Recent volcanoes are:

Fig. 2.4. AFM-diagram after Rickwood (1989) of sourthern Peru data

Rare earth elements (REE)

Description of REE pattern

Eocene and Miocene volcanoes

Mafic volcanoes (monogenetic)

2.4.1.1 Temporal variations of element ratios

Fig. 2.13. REE ratios of lavas and ignimbrites of southern Peru.

Spatial and Temporal variations of Sr-Nd isotopes

Fig. 2.16. 87Sr/86Sr and 143Nd/144Nd ratios versus Age.

Spatial and Temporal variations of Pb-isotopes

Fig. 2.20. SiO2 versus La/Ta, La/Yb and Sr/Y.

3 Regional and temporal patterns in Meso-Cenozoic magmatic evolution in the Central

3.3 Geochemistry of Meso-Cenozoic magmatic rocks in the CA

3.3.1 Major and trace elements

Rare earth elements (REE)

which are transformed to:

Spatial and Temporal variation of Sr-Nd isotopes

Fig. 3.10. 87Sr/86Sr and εNd ratios versus Latitude °S.

Spatial and Temporal variation of Pb- isotopes

Fig. 3.13. Regional variation in Pb isotope compositions along the CA.

3.4 Summary and Discussion: Implications for a tectonic model

• (87Sr/86Sr)initial ratios increase from Jurassic until present.