Dissertation Philipp A. Brandl

The geochemistry of oceanic basalts:
constraints on melting and
composition of the Earth’s mantle
Die Geochemie ozeanischer Basalte:
Einblicke in Schmelzprozesse und
Zusammensetzung des Erdmantels
Der Naturwissenschaftlichen Fakultät
der Friedrich-Alexander-Universität
Erlangen-Nürnberg
zur
Erlangung des Doktorgrades Dr. rer. nat.
vorgelegt von
Philipp A. Brandl
aus Hirschau/Opf.
Als Dissertation genehmigt
von der Naturwissenschaftlichen Fakultät
der Friedrich-Alexander-Universität Erlangen-Nürnberg
Tag der mündlichen Prüfung: 29.05.2013
Vorsitzender der Promotionskommission: Prof. Dr. J. A. C. Barth
Erstberichterstatter/in: Prof. Dr. K. M. Haase
Zweitberichterstatter/in: Prof. Dr. R. Klemd
Wissen und Erkennen sind die Freude und die Berechtigung der Menschheit.“
”
Alexander von Humboldt (1769–1859)
Philipp A. Brandl
www.researcherid.com/rid/F-5576-2012
GeoZentrum Nordbayern
Department für Geographie und Geowissenschaften
Friedrich-Alexander-Universität Erlangen-Nürnberg
Erlangen, 2013
edited with LATEX
Abstract
Magmatic activity in the ocean basins is the largest contributor to the volume of global
magmatism. Most oceanic igneous activity is focused along the mid-ocean ridges. The
mid-ocean ridge systems spans more than 60,000 km around the globe and is one of the
primary geodynamic settings that allows to investigate the exchange of elements and heat
between the Earth’s interior and its surface. The geodynamic changes from a steady state
along the mid-ocean ridges may thus influence the environmental conditions on Earth
such as plate tectonics, eustatic sealevel (as a function of ridge depth) and climate (by
volcanic degassing). Detailed studies of oceanic volcanism and magmatism in general are
thus required to better constrain global processes and elemental cycles.
This thesis focuses on three major aspects that, in combination, significantly enhance
our understanding of eruptive processes at mid-ocean ridges, melting in the mantle and
mantle evolution over extended periods of time of 10–100 Ma. I used the major element
composition of fresh mid-ocean ridge basalt (MORB) glasses to reconstruct the mantle
potential temperature at which the parental magmas formed (chapter 3). The samples
used have been obtained from old seafloor (6–170 Ma) through ocean drilling and thus
allow to infer on the thermal evolution of the mantle since the Jurassic. The results of this
study indicate that mantle temperatures remain higher under supercontinents (by continental insulation) and MORB erupted immediately after breakup record 50–150◦ C higher
mantle temperatures compared to values several million years after the initial opening of
an ocean.
Current models of the melting processes in the mantle are still insufficient to allow a precise
quantitative estimate of the mantle composition and the physical conditions of melting.
My studies on Seamount 6 (chapter 4) and the extinct spreading centre of the Galapagos
Rise (chapter 5) contribute to our understanding of melting at or near mid-ocean ridges.
Geochemical analyses of samples from these two locations preserve an extreme chemical
variability, indicating a large compositional variability of their mantle source. State-ofthe-art melting models give further insights into the melting behaviour of depleted and
enriched mantle material. Enriched melts form deeper, but become progressively more
diluted at higher degrees of partial melting in melts from the ambient depleted mantle.
As a result, the geochemical signatures of mantle heterogeneity are increasingly diluted
Dissertation P.A. Brandl
i
at high degrees of partial melting (e.g., atmid-ocean ridges) and the composition of the
depleted mantle may therefore be more depleted than previously thought.
The third aspect of my thesis deals with the accretion of oceanic crust along mid-ocean
ridges in order to constrain its precise composition and internal structure. The study at
the southern Mid-Atlantic Ridge (chapter 6) reveals insights into accretionary processes
and volcanostratigraphy of the oceanic crust at slow spreading rates (20–55 mm a−1 full
spreading rate). The eruption of geochemically distinct melts within only few kilometres
distance indicates that the mantle even in these slow spreading regions is highly heterogeneous and that melts underneath the ridge axis rise in small, chemically isolated batches.
Eruptive stages are interrupted by phases of amagmatic, tectonic activity. Oceanic crust
at slow-spreading centres is in consquence heterogeneous not only in terms of its structure
but also chemical composition.
The primary conclusions of this thesis with respect to the geochemical interpretation
of oceanic basalts are: a) supercontinents and continental breakup have a major impact
on temperature and convection in the mantle, b) oceanic basalts represent complex mixtures of melts of a heterogeneous mantle source and their geochemical interpretation is
not straightforward and c) oceanic crust formed at slow-spreading ridges is composed
of highly variable crustal units formed by the eruption of chemically isolated batches of
magma.
ii
Dissertation P.A. Brandl
Kurzfassung
Ein Großteil der magmatischen Aktivität auf unserer Erde findet in den ozeanischen
Becken statt, vor allem entlang der Mittelozeanischen Rücken. Diese umspannen den
Globus auf einer Länge von 60.000 km und stellen eine der wichtigsten Schnittstellen für
den Austausch von Elementen und Hitze zwischen dem Erdinneren und der Oberfläche
dar. Geodynamische Änderungen entlang der Mittelozeanischen Rücken beeinflussen möglicherweise auch die Umweltbedingungen auf der Erde, wie beispielsweise die Plattentektonik, eustatische Meeresspiegelschwankungen (als Funktion der Rückentiefe) und Klima
(durch vulkanische Entgasung). Genaue Untersuchungen zum Verständnis ozeanischen
Vulkanimus und Magmatismus sind daher wichtig, um die globalen Prozesse und Stoffkreisläufe besser zu verstehen.
Diese Doktorarbeit behandelt im Wesentlichen drei Hauptaspekte, die zusammengenommen zu unserem Verständnis der Prozesse an Mittelozeanischen Rücken, Schmelzentstehung im Mantel und Mantelentwicklung über Zeitskalen von 10–100 Millionen Jahren
beitragen. Die Hauptelementzusammensetzung frischer, glasiger mittelozeanischer Rückenbasalte (MORB) kann dazu benutzt werden, die Temperaturen zu rekonstruieren, die im
Erdmantel während der Schmelzbildung herrschten. Meine Proben, die von altem Ozeanboden (6–170 Mio. Jahre) stammen und durch Bohrungen gewonnen wurden, ermöglichen
es, die thermische Entwicklung des Erdmantels seit dem Jura zu untersuchen. Die Ergebnisse dieses Projekts (Kapitel 3) weisen darauf hin, dass es unter Superkontinenten zu
einem Aufstau der Hitze kommt, die auf eine Isolierung des Erdmantels durch die
aufliegende, kontinentale Lithosphäre verursacht wird. In der Folge sind die Temperaturen des Erdmantels nach dem Auseinanderbrechen des Kontinents 50–150◦ C heißer als
dies heute, nach mehreren Jahrmillionen der Ozeanbodenspreizung, an Mittelozeanischen
Rücken beobachtet wird.
Die Aufschmelzprozesse im Erdmantel sind nach wie vor nicht gut genug untersucht,
um genaue Rückschlüsse auf die Zusammensetzung des Erdmantels und die physikalischen Bedingungen während der Aufschmelzung zu ziehen. Meine Arbeiten am submarinen Vulkan Seamount 6“(Kapitel 4) und der erloschenen Spreizungsachse des Galapa”
”
gos Rise“(Kapitel 5) tragen zu unserem Verständnis der Schmelzbildung an oder nahe
Mittelozeanischer Rücken bei. Die geochemische Analysen der Gesteine dieser beiden
Dissertation P.A. Brandl
iii
Lokalitäten spiegeln eine extreme chemische Bandbreite wieder, die auf eine große chemische Variabilität des Erdmantels hinweisen. Die Anwendung moderner Schmelzmodellierungen ermöglicht weitere Einblicke in das Aufschmelzverhalten chemisch verarmter
und angereicherter Mantelquellen. Chemisch angereicherte Schmelzen entstehen bereits in größerer Tiefe werden aber bei steigenden Aufschmelzgraden zunehmend durch
die Schmelze des verarmten oberen Mantels verdünnt. Als Folge geht die geochemische
Signatur des angereicherten Mantels bei höheren Aufschmelzgraden (wie z.B. an Mittelozeanischen Rücken) verloren und es ist möglich, dass der verarmte Mantel chemisch
noch stärker verarmt ist als dies bisher vermutet wurde.
Der dritte Teil meiner Doktorarbeit befasst sich mit dem Entstehungsprozess ozeanischer Kruste, um mehr über deren genaue Zusammensetzung und Struktur zu erfahren.
Die Untersuchungen am südlichen Mittelatlantischen Rücken (Kapitel 6) gewähren Einblick in den Akkretionsprozess und die vulkanische Stratigraphie ozeanischer Kruste an
langsam spreizenden Achsen (volle Spreizungsrate: 20–55 mm a−1 ). Die Eruption chemisch
stark unterschiedlicher Lavaeinheiten, in nur wenigen Kilometern Entfernung voneinander
weist auf einen stark heterogenen Erdmantel hin. Sie zeigen, dass Schmelzen unter der
Rückenachse in kleinen, chemisch isolierten Körpern aufsteigen. Unterbrochen werden die
einzelnen Eruptivphasen von Zeitabschnitten, in denen überwiegend amagmatische, tektonische Aktivität statt findet. Als Folge ist die ozeanische Kruste langsam spreizender
Rücken nicht nur im Bezug auf die interne Struktur, sondern auch in ihrer chemischen
Zusammensetzung stark variabel.
Die Hauptergebnisse meiner Arbeit im Hinblick auf die Geochemie ozeanischer Basalte
sind: a) Superkontinente und deren Auseinanderbrechen haben einen bedeutenden Einfluss auf die Mantelkonvektion und -temperatur, b) ozeanische Basalte sind komplexe
Mischungen von Schmelzen eines heterogenen Erdmantels mit der Folge, dass deren geochemische Interpretation sehr komplex ist, sowie c) ozeanische Kruste, die sich an langsam
spreizender Rücken gebildet hat, besteht aus verschiedenen Krusteneinheiten, die durch
die Eruption chemisch isolierter, heterogener Magmenkörper gebildet werden.
iv
Dissertation P.A. Brandl
Statement of candidate
I certify that the work in this thesis entitled “The geochemistry of oceanic basalts:
constraints on melting and composition of the Earth’s mantle” has previously
not been submitted for any degree nor has it been submitted as part of requirements
for a degree to any other university or institution other than the Friedrich-AlexanderUniversität Erlangen-Nürnberg.
I also certify that the thesis is a new, original piece of research and it has been written by me. Any help and assistance that I have received in my research work and the
preparation of the thesis itself have been appropriately acknowledged.
In addition, I certify that all information sources and literature used are indicated in
the thesis.
This thesis contains material that has been published or accepted for publication in peerreviewed ISI-journals or is in preparation for publication, as follows:
Chapter 3 “High mantle temperatures following rifting caused by continental insulation” has been accepted for publication in Nature Geoscience (accepted: 07.02.2013)
and is currently in press. My contribution to this publication consisted of sampling, data
analyses and interpretation and parts of writing the text, resulting in a total contribution
of about 70%. Impact factor: 11.754 (2011).
Chapter 4 “Volcanism on the flanks of the East Pacific Rise: quantitative constraints on mantle heterogeneity and melting processes” has been published in
Chemical Geology in 2012. My contribution to this publication consisted of parts of analytical work (major elements and radiogenic isotopes), data interpretation, modelling and
parts of writing the text, resulting in a total contribution of about 75%. Impact factor:
3.518 (2011). doi:10.1016/j.chemgeo.2011.12.015.
Chapter 5 “Insights into mantle composition and mantle melting beneath midocean ridges from postspreading volcanism on the fossil Galapagos Rise” has
been published in Geochemistry Geophysics Geosystems in 2011. My contribution to this
Dissertation P.A. Brandl
v
publication consisted of modelling and minor parts of writing the text, resulting in a total
contribution of about 15%. Impact factor: 3.021 (2011). doi:10.1029/2010GC003482. An
edited version of this paper was published by AGU. Copyright 2011 American Geophysical Union.
Chapter 6 “Compositional variation of lavas from a young volcanic field on
the Southern Mid-Atlantic Ridge, 8◦ 48’S” is currently in preparation for publication. My contribution to this publication consisted of geologic interpretation of cruise and
submersible data, the production of maps and parts of writing the text, resulting in a
total contribution of about 40%.
Erlangen, 13.02.2013
Philipp A. Brandl
vi
Dissertation P.A. Brandl
Full publication list
Peer-reviewed publications
Brandl, P. A., Regelous, M., Beier, C. & Haase, K. M. (in press): Continental insulation
and the thermal evolution of the upper mantle. Nature Geoscience, accepted: 07.02.2013.
Genske, F. S., Beier, C., Haase, K. M., Turner, S. P., Krumm, S. & Brandl, P. A. (in
press): Oxygen isotopes in the Azores islands: Crustal assimilation recorded in olivine.
Geology, accepted: 08.11.2012.
Lehnert, O., Stouge, S. & Brandl, P. A. (in press): Conodont biostratigraphy in the Early
to Middle Ordovician strata of the Oslobreen Group in Ny Friesland, Svalbard. ZDGG
(German Journal of Geosciences), available online: 29.01.2013, doi: 10.1127/1860-1804/
2013/0003.
Beier, C., Mata, J., Stöckhert, F., Mattielli, N., Brandl, P. A., Madureira, P., Genske, F.
S., Martins, S., Madeira, J. & Haase, K. M. (in press): Geochemical evidence for melting
of carbonated peridotite on Santa Maria Island, Azores. Contributions to Mineralogy and
Petrology, available online: 07.12.2012, doi: 10.1007/s00410-012-0837-2.
Brandl, P. A., Beier, C., Regelous, M., Abouchami, W., Haase, K. M., Garbe-Schönberg, D.
& Galer, S. J. G. (2012): Volcanism on the flanks of the East Pacific Rise: quantitative
constraints on mantle heterogeneity and melting processes. Chemical Geology 289-299
(3-4), 41-56, doi: 10.1016/j.chemgeo.2011.12.015.
Haase, K. M., Regelous, M., Duncan, R. A., Brandl, P. A., Stroncik, N. & Grevemeyer, I.
(2011): Insights into mantle composition and mantle melting beneath mid-ocean ridges
from post-spreading volcanism on the fossil Galapagos Rise. G3 - Geochemistry Geophysics
Geosystems 12, Q0AC11, doi: 10.1029/2010GC003482.
Dissertation P.A. Brandl
vii
Conference abstracts
Brandl, P. A., Genske, F. S., Haase, K. M. & Beier, C. (2013): Quaternary volcanism
in Central Europe: new results from Železná Hůrka/Eisenbühl (Czech Republic). Basalt
2013, Görlitz.
Brandl, P. A., Regelous, M., Beier, C. & Haase, K .M. (2013): High mantle temperatures recorded in post-breakup MORB. IODP-ICDP Meeting, Freiberg.
Brandl, P. A., Regelous, M., Beier, C. & Haase, K. M. (2012): Continental breakup and
its effect on MORB chemistry. AGU Fall Meeting Abstract Volume, T33G-2736.
Haase, K. M., Brandl, P. A., Melchert, B., Hauff, F., Garbe-Schönberg, D., Paulick, H.,
Kokfelt, T. H. & Devey, C. W. (2012): Compositional variation of lavas from a young
volcanic field on the Southern Mid-Atlantic Ridge, 8o 48’S. AGU Fall Meeting Abstract
Volume, V11D-2806.
Brandl, P. A., Regelous, M., Beier, C. & Haase, K. M. (2012): Mesozoic MORB. 22nd
V.M. Goldschmidt conference, Montréal.
Brandl, P. A., Regelous, M., Haase, K. M. & Beier, C. (2012): Tracing the evolution of
the upper mantle using ancient MORB glasses. IODP-ICDP Meeting, Kiel.
Beier, C., Nichols, A. R. L., Brandl, P. A., Brätz, H. & Expedition 330 Scientists (2012):
Geochemical Constraints on the evolution of the Louisville Seamount Trail. IODP-ICDP
Meeting, Kiel.
Brandl, P. A., Regelous, M., Beier, C., & Haase, K. M. (2011): Chemical Evolution of
the Oceanic Crust on 103 –108 Year Timescales. AGU Fall Meeting Abstract Volume,
DI13A-2149.
Brandl, P. A., Regelous, M. & Haase, K. M. (2011): The evolution of the Earth’s mantle:
new insights from old seafloor. 21st V.M. Goldschmidt conference, Prague.
Regelous, M., Haase, K. M. & Brandl, P. A. (2011): Oceanic basalts provide a biased
view of mantle composition. 21st V.M. Goldschmidt conference, Prague.
Brandl, P. A., Lehnert, O. & Stouge, S. (2010): Conodonts, isotopes, sea-level & plate
tectonics - the origin of NE Spitsbergen in the peri-Laurentian terrane puzzle. CASE
viii
Dissertation P.A. Brandl
meeting, BGR, Hannover.
Brandl, P. A., Beier, C., Regelous, M., Haase, K. M., Abouchami, W. & Garbe-Schönberg, D. (2010): Off-axis seamounts along the East Pacific Rise - inferences on melt extraction and source heterogeneities in the upper mantle. 88th annual conference of Deutsche
Mineralogische Gesellschaft (DMG), Münster.
Lehnert, O., Stouge, S. & Brandl, P. A. (2009): Conodont faunas and carbon isotopes of
the Oslobreen Group, Ny Friesland (NE Spitsbergen): correlation along the Laurentian
margin. Absolutely final meeting of IGCP 503:“Ordovician palaeogeography and palaeoclimate”, Copenhagen.
Lehnert, O., Stouge, S. & Brandl, P. A. (2009): The Tremadocian through Darriwilian
conodont succession of NE Spitsbergen: faunal affinities and intercontinental correlation.
Paleozoic Seas Symposium, Graz.
Dissertation P.A. Brandl
ix
Acknowledgments
First of all, I would like to thank Prof. Dr. Karsten M. Haase and Dr. Marcel Regelous
for supervising me and giving me the opportunity to work on this great project. I am
very greatful to Dr. Christoph Beier for his patience during countless discussions, proofreading, coffee and sport breaks and company on the institute’s hoppy balcony.
I would like to thank my colleagues Dr. Wafa Abouchami, Dr. Helene Brätz, Dr. Stephen
J. G. Galer, Dr. Dieter Garbe-Schönberg, Prof. apl. Dr. Michael M. Joachimski, Prof. Dr.
Reiner Klemd, Dr. Oliver Lehnert, Dr. John Maclennan and Prof. Dr. Andreas Stracke
for scientific cooperations, inspiring discussions and support.
I would like also to thank all my other colleagues at the GeoZentrum Nordbayern and
friends around the globe for sharing their knowledge and, much more important, great moments and funny evenings. These are by name: Judith Beier, Heidi Daxberger,
Sebastian Dittrich, Sarah Freund, Andrea Friese, Dr. Felix S. Genske, Manuel Keith,
Dr. Stefan H. Krumm, Melanie Meyer (especially for helping with the quirks of LATEX and
proof-reading), Amir Mohammadi, Volker Möller, Lukas Pflug, Andreas Richter, Ludwig
Ritschl, Isabell Schiemer, Henning Schulz, Christoph Weinzierl, and Manuel Winkler.
I am grateful to the technical staff of the GeoZentrum Nordbayern for their great support
during my studies. These include namely Melanie Hertel, Veronika Kühnert, Daniele Lutz,
Konrad Kunz, Christine Scharf and Bernd Schleifer.
I deeply acknowledge the great support of my parents, Jutta and Friedrich Brandl, during
school, university and at any other time. Thanks also to my brother Ferdinand Brandl
and my sister Ina Meillan and her family.
Finally, I would like to thank Deutsche Forschungsgemeinschaft (DFG) and the Erika
Gierhl Foundation for providing funding. I apologise in advance in case that somebody is
missing here. I am solely responsible for any remaining errors and omissions of this work.
x
Dissertation P.A. Brandl
Contents
Contents
Abstract
i
Kurzfassung
iii
Statement of candidate
v
Full publication list
vii
Acknowledgements
x
Contents
xi
List of Figures
xiv
List of Tables
xvi
1 Introduction
1.1 Evolution of the Earth’s mantle . . . . . . . . . .
1.1.1 Origin of the geochemical mantle reservoirs
1.1.2 The thermal history of the mantle . . . . .
1.2 Geochemistry of oceanic basalts . . . . . . . . . .
1.2.1 Melting in the mantle . . . . . . . . . . . .
1.2.2 Melt extraction and magma mixing . . . .
1.2.3 Shallow-level processes . . . . . . . . . . .
1.3 Structure of the oceanic igneous crust . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
. 1
. 1
. 3
. 5
. 5
. 8
. 10
. 12
2 Aims of the study
15
2.1 Constraining the thermal evolution of the upper mantle . . . . . . . . . . . 15
2.2 Insights into mantle composition and melting . . . . . . . . . . . . . . . . 16
2.3 Studying volcanic processes at mid-ocean ridges . . . . . . . . . . . . . . . 17
3 High mantle temperatures following rifting
3.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4 Results & Discussion . . . . . . . . . . . . . . . . . . . . . .
3.5 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .
3.6 Supplementary Information . . . . . . . . . . . . . . . . . .
3.6.1 Sampling and analytical methods . . . . . . . . . . .
3.6.2 Zero-age MORB reference database . . . . . . . . . .
3.6.3 Correction for the effects of fractional crystallisation .
3.6.4 Calculation of primary magma compositions . . . . .
Dissertation P.A. Brandl
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
19
19
20
20
21
26
27
27
28
31
31
xi
Contents
3.6.5
3.6.6
3.6.7
Comparison of ancient and zero-age MORB datasets . . . . . . . . 33
Calculation of mantle potential temperatures . . . . . . . . . . . . . 37
Effects of mantle heterogeneity and hotspots . . . . . . . . . . . . . 39
4 Volcanism on the flanks of the East Pacific Rise
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 Geological setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 Samples and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.1 Petrography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.2 Major and trace element composition . . . . . . . . . . . . . . . .
4.4.3 Radiogenic isotopes . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5.1 The effects of fractional crystallisation . . . . . . . . . . . . . . .
4.5.2 The role of mixing processes . . . . . . . . . . . . . . . . . . . . .
4.5.3 Highly heterogeneous mantle beneath Seamount 6 . . . . . . . . .
4.5.4 Melting of a heterogeneous mantle . . . . . . . . . . . . . . . . .
Melting models and input parameters . . . . . . . . . . . . . . . .
Modelling results and implications for seamount lava petrogenesis
4.5.5 Implications for the use of oceanic lavas as probes of mantle composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6 Supplementary Information . . . . . . . . . . . . . . . . . . . . . . . . .
5 Post-spreading volcanism on the fossil Galapagos Rise
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Tectonic setting and sample locations . . . . . . . . . . . . . . . . . . .
5.3 Samples and analytical methods . . . . . . . . . . . . . . . . . . . . . .
5.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.4.1 Ar-Ar ages of Galapagos Rise lavas . . . . . . . . . . . . . . . .
5.4.2 Major and trace element geochemistry . . . . . . . . . . . . . .
5.4.3 Sr, Nd, and Pb isotope compositions . . . . . . . . . . . . . . .
5.4.4 Comparison with lavas from other extinct spreading centres . .
5.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.5.1 Origin of chemical and isotopic variations . . . . . . . . . . . . .
Effects of fractional crystallisation and melting processes . . . .
The role of mantle heterogeneity . . . . . . . . . . . . . . . . . .
Melting a two-component mantle . . . . . . . . . . . . . . . . .
5.5.2 Mantle upwelling and melting beneath spreading ridges . . . . .
5.5.3 Implications for chemical and isotopic variation in global MORB
5.6 Summary and conclusion . . . . . . . . . . . . . . . . . . . . . . . . . .
6 Compositional variation of lavas from a young volcanic field
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2 Geological setting . . . . . . . . . . . . . . . . . . . . . . .
6.3 Sampling and analytical methods . . . . . . . . . . . . . .
6.3.1 Sampling and observations . . . . . . . . . . . . . .
6.3.2 Geochemical analyses of glasses . . . . . . . . . . .
6.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xii
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
41
42
44
46
48
48
48
52
53
53
55
57
58
58
60
. 63
. 65
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
69
70
72
74
75
75
76
79
80
82
82
82
84
84
87
89
91
.
.
.
.
.
.
93
94
96
96
96
98
98
Dissertation P.A. Brandl
Contents
6.5
6.6
6.4.1 Geological observations on the volcanic field . . . . . .
6.4.2 Petrography of the lavas . . . . . . . . . . . . . . . . .
6.4.3 Composition of the volcanic glasses . . . . . . . . . . .
6.4.4 Along axis variations of compositions . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.5.1 Definition and formation of the lava flow units . . . . .
6.5.2 Composition and petrogenesis of the southern lava unit
6.5.3 Constraints on eruption ages . . . . . . . . . . . . . . .
6.5.4 Magma ascent beneath the slow-spreading A2 segment
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
98
101
101
103
104
104
105
106
107
108
7 Synthesis & Outlook
109
7.1 The evolution of the upper mantle . . . . . . . . . . . . . . . . . . . . . . . 109
7.2 Constraints on mantle melting and mixing processes . . . . . . . . . . . . . 110
7.3 Volcanic eruptions at mid-ocean ridges . . . . . . . . . . . . . . . . . . . . 111
References
113
Appendix
131
Dissertation P.A. Brandl
xiii
List of Figures
List of Figures
xiv
1.1
1.2
1.3
1.4
1.5
1.6
1.7
Schematic model of oceanic volcanism . . . . . . . .
Geochemical mantle reservoirs . . . . . . . . . . . .
Heatflow to the Earth’s surface . . . . . . . . . . .
Anhydrous P-T diagram for mantle lherzolite . . .
Schematic model of the magmatic plumbing system
Cross-section of the oceanic crust . . . . . . . . . .
Ridge-type morphology . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
Fractionation-corrected major element compositions . . . . . . . . . . . . .
Temporal variations in composition of MORB . . . . . . . . . . . . . . . .
Major element compositions of MORB . . . . . . . . . . . . . . . . . . . .
Evolution of mantle temperature with time following continental breakup. .
Global map of the age of the oceanic crust . . . . . . . . . . . . . . . . . .
Major element composition of MORB . . . . . . . . . . . . . . . . . . . . .
Na8 versus Fe8 of ancient and zero-age MORB . . . . . . . . . . . . . . . .
MgO versus FeOT and CaO for ancient MORB glasses . . . . . . . . . . .
Frequency of sampled spreading ridge water depths . . . . . . . . . . . . .
Variation of Na8 and Fe8 with spreading ridge depth . . . . . . . . . . . . .
Comparison of data of modern MORB and ancient MORB . . . . . . . . .
Age of oceanic crust versus penetration into the igneous oceanic crust . . .
Pacific off-axis MORB . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mantle potential temperature of zero-age MORB . . . . . . . . . . . . . .
Fractionation corrected Na8 and Fe8 values for zero-age MORB . . . . . . .
Chondrite-normalised (La/Sm)N and age (Ma) . . . . . . . . . . . . . . . .
22
23
24
25
27
29
30
32
34
34
35
36
37
38
39
40
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
Bathymetric map of Seamount 6. . . . . . . . . . . . . . .
Cross-section through Seamount 6 . . . . . . . . . . . . . .
Major element compositions of Seamount 6 lavas . . . . . .
Trace element concentrations of Seamount 6 lavas . . . . .
Zr and Nb concentrations in basalts and andesites . . . . .
Sr, Nd and Pb isotope composition of lavas from Seamount
Chemical composition of mixing endmembers . . . . . . .
Results of a melting model . . . . . . . . . . . . . . . . . .
5.1
5.2
5.3
5.4
5.5
5.6
5.7
Tectonic map of the eastern Pacific . . . . . . . . . . . . . .
Bathymetric map of the Galapagos Rise . . . . . . . . . . .
Major element compositions of Galapagos Rise lavas. . . . .
Trace element compositions of Galapagos Rise lavas . . . . .
Radiogenic isotope compositions of lavas from the Galapagos
Trace element and isotope compositions of lavas . . . . . . .
Comparison of trace element and isotope compositions . . .
.
.
.
.
.
6
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 2
. 3
. 4
. 6
. 9
. 12
. 13
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
45
46
50
51
52
53
56
61
. . .
. . .
. . .
. . .
Rise
. . .
. . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
72
73
77
78
80
81
82
Dissertation P.A. Brandl
List of Figures
5.8
Results from modelling two-component mixing of melts . . . . . . . . . . . 85
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
Bathymetric map of the Southern Mid-Atlantic Ridge . .
Detailed maps of the young volcanic field . . . . . . . . .
Summary of ROV dive 43 . . . . . . . . . . . . . . . . .
Photographs of representative lavas from the ROV dives
Major element diagrams of glass samples . . . . . . . . .
Chemical variation with latitude . . . . . . . . . . . . . .
Partial melting vs. source composition . . . . . . . . . .
(226 Ra/230 Th) versus Ba/Th for the A2 lavas . . . . . . .
Dissertation P.A. Brandl
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
95
97
99
100
102
103
104
106
xv
List of Tables
List of Tables
xvi
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
Lava flow type, alteration and petrography of Seamount 6 samples. . .
Linear least-squares regression parameters for Pb isotope data‡ . . . . .
Input parameters used in the melting models illustrated in Figure 4.8∗ .
Partition coefficients used in Stracke and Bourdon (2009) . . . . . . . .
Mineral mode: Lherzolite (peridotite) . . . . . . . . . . . . . . . . . . .
Melting mode: Lherzolite (peridotite) . . . . . . . . . . . . . . . . . . .
Spinel-garnet transition: sign reversed to the rest of the reaction . . . .
Mineral mode: Pyroxenite . . . . . . . . . . . . . . . . . . . . . . . . .
Melting mode: Pyroxenite . . . . . . . . . . . . . . . . . . . . . . . . .
Input source composition . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
49
57
59
65
66
66
66
66
66
67
5.1
Summary of 40 Ar/39 Ar data for lavas from the Galapagos Rise Fossil Spreading Centre.† . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Dissertation P.A. Brandl
1. Introduction
1 Introduction
1.1 Evolution of the Earth’s mantle
1.1.1 Origin of the geochemical mantle reservoirs
The age of the Earth is deciphered to be about 4.51–4.55 Ga (e.g., Manhes et al., 1980;
Dalrymple, 2001 and references therein). Nevertheless, the process of chemical differentiation was not finished by that time but continues to the present. The core, mantle, crust
(oceanic and continental crust) and atmosphere formed by chemical differentiation, particularly by partial melting and segregation of these melts from their mantle source (e.g.,
DePaolo and Wasserburg, 1976; Allègre, 1982; Zindler and Hart, 1986; Hofmann, 1988 and
further discussion below). As a result, large portions of the Earth’s mantle are depleted
in those elements that are nowadays concentrated in the crust and the atmosphere. The
continental crust, for example, is 40–80 times more enriched in the most incompatible elements (Cs, Rb, K, Ba, Th, U and La) compared to the depleted mantle (e.g., Hofmann,
1988; Taylor and McLennan, 1995). The overall proportion of the Primitive Mantle (PM)
that has undergone depletion by partial melting is estimated to range somewhere between
30 and 100% (Stracke, 2008). However, not only the proportion, also the precise timing (or
duration) of depletion in the Earth’s mantle is not yet well constrained. Since the depletion of the mantle is mainly controlled by continental crust extraction, the main controls
on mantle depletion are volume and rate of continental growth (e.g., Allègre et al., 1980;
Workman and Hart, 2005). With respect to these observations it is not conclusively determined whether the process of continental growth is continuous (Hurley and Rand, 1969),
episodic (e.g., McCulloch and Bennett, 1994; Taylor and McLennan, 1995; Condie, 2000)
or was completed prior to about 3 Ga and since then, continental crust has only been
recycled (e.g., Armstrong, 1968; Fyfe, 1978).
However, the overall result of differentiation and recycling processes is a heterogeneous
mantle that evolved into numerous chemically distinct reservoirs (Fig. 1.1), especially in
terms of radiogenic (e.g., 87 Sr/86 Sr, 143 Nd/144 Nd, 206 Pb/204 Pb) isotopes. Gast et al. (1964)
were the first to recognise that plume-related basalts (Ascension and Gough Islands) are
different in their strontium and lead isotope composition compared to MORB. Lead isotope studies (e.g., Hofmann and White, 1982; White and Hofmann, 1982) argued that
Dissertation P.A. Brandl
1
1. Introduction
Figure 1.1: Schematic model of oceanic volcanism (modified from Winter, 2001). The different
geodynamic settings (e.g., mid-ocean ridges, ocean island) also have distinct mantle sources. The upper
mantle is variably Depleted ;antle (DM), whereas the lower mantle contains enriched mantle material (e.g.,
PREMA, HIMU). For further discussion see main text. Nomenclature from Zindler and Hart (1986).
recycling of oceanic crust is the most likely process that accounts for the observed isotopic
heterogeneity in the mantle. Thus, the structure of the Earth’s mantle is often considered
as a ‘marble cake mantle’ with mantle heterogeneities of variable size and distribution
(Allègre and Turcotte, 1986). According to Zindler and Hart (1986) these heterogeneities
comprise four ‘common’ endmembers: HIMU (high-µ; µ=238 U/204 Pb), EM-1 (Enriched
Mantle 1), EM-2 (Enriched Mantle 2) and DMM (Depleted MORB Mantle) with the two
additional components PUM (Primitive Upper Mantle) and PREMA (Prevalent Mantle
synonymous to ‘FOZO’ or ‘C’; Stracke et al., 2005). Generally, DMM has the least enriched isotope compositions in Sr-Nd-Pb(-Hf) spaces (as a result of depletion by partial
melting and the extraction of incompatible elements such as Rb, Th, U), whereas HIMU
is characterised by the most radiogenic 206 Pb/204 Pb (Fig. 1.2) ratios of all reservoirs, resulting from recycling of igneous oceanic crust and ‘aging’ of the isotopic composition
over 0.5–3 Ga (Stracke, 2003 and references therein). The enriched mantle endmembers
EM-1 and EM-2 are generally characterised by low 143 Nd/144 Nd and high 207 Pb/204 Pb and
208
Pb/204 Pb at a given 206 Pb/204 Pb (Zindler and Hart, 1986). EM-1 has the most radiogenic 177 Hf/176 Hf (Stracke et al., 2003), whereas EM-2 has the most radiogenic 87 Sr/86 Sr
(Fig. 1.2; Zindler and Hart, 1986). Thus, the origin of these two enriched mantle reservoirs
is commonly interpreted to have formed by recycling of ‘pelagic’ (EM-1) or ‘terrigeneous’
(EM-2) sediments (see Zindler and Hart, 1986; Stracke et al., 2003 and references therein).
The mass proportions of these heterogeneities are in the order of ∼43% PM, ∼33% DMM
and 5–21% Ocean Island Basalt (OIB) reservoirs (HIMU, EM-1 and EM-2; Workman and
Hart, 2005).
2
Dissertation P.A. Brandl
1. Introduction
Figure 1.2: Representative
Srand Pb-isotope data for the four
‘common’ mantle endmembers
of Zindler and Hart (1986). a)
208
Pb/204 Pb versus 206 Pb/204 Pb b)
206
Pb/204 Pb versus 87 Sr/86 Sr of representative samples from ocean islands
(e.g., Cook-Austral islands, Hawaii)
and MORB. Precise information on
data sources can be found in Hofmann
(2003). Figure modified from Hofmann
(2003).
1.1.2 The thermal history of the mantle
As discussed above, the DMM reservoir originates from PM through partial melting and
melt extraction. The total amount of mass removed from the primitive mantle to produce
the DMM is estimated to be about 2–3% (McDonough and Sun, 1995). However, this
relatively small proportion results in a decrease in the radiogenic heat production to only
10–30% of that in the PM (Jochum et al., 1983), which has important implications for
the nature of mantle convection.
The energy budget of the Earth is commonly expressed by the ratio of heat flow to
heat production (‘Urey’-ratio; Zindler and Hart, 1986). Heat in the Earth’s interior is
predominantly produced by the radioactive decay of 238 U, 234 U, 232 Th and 40 K in the
mantle (e.g., Urey, 1956; Wasserburg et al., 1964). However, it is not conclusively clear to
which amount the core is contributing to the total heat flow (one silicate earth K budget
versus fossil heat production, resulting in a contribution of 15–60%; Zindler and Hart,
1986) and whether the whole-mantle or layered-mantle convection hypothesis might be
more realistic (see discussion in Zindler and Hart, 1986). The heat contrast between heat
production in the Earth’s interior (core and PM) and heat loss (cooling) to space must
be balanced either by conductive or convective heat transport. Richter (1984) applied a
Dissertation P.A. Brandl
3
1. Introduction
regionalised heat flow model and concluded that the heat flow in oceanic regions is at least
3 to 4 times greater than in continental regions (Fig. 1.3). Heat transport through the
‘conductive lid’ (Richter, 1984) of continents and continental shelves is thus significantly
less efficient compared to heat transport by convective processes such as the formation of
new crust and orogeny (Sclater et al., 1980). Mid-ocean ridge processes (where new plates
are created) are thus important for the global heat flow regime and mantle convection.
Figure 1.3: Heatflow to the Earth’s surface. The rate of heatflow at mid-ocean ridges (white) is
generally higher by almost one magnitude compared to the continents. Figure modified from Jaupart and
Mareschal (2010).
As discussed above, the thermal state of the mantle is controlled by the global Urey-ratio
and energy (heat) contrast within the mantle. Since the chemical differentiation of the
mantle is a continuous process and has influence on the capability of heat production
by radioactive decay, one might expect that the mantle (potential) temperature varies
through the Earth’s history. Oceanic lavas (such as MORB and OIB) record the thermal state of the mantle through their petrology and geochemical composition (Herzberg
et al., 2007). Thus, these rocks can help to reconstruct the thermal history of the mantle
through time and indeed, Al-undepleted komatiites (with about 27–30% MgO) indicate
higher mantle temperatures in the Archean compared to the present-day mantle (Herzberg
et al., 2007).
The thermal evolution of the Earth’s mantle is thus characterised by a gradual (secular)
4
Dissertation P.A. Brandl
1. Introduction
cooling of the ambient mantle. Estimates for the decrease in mantle potential temperature since the Archean are on the order of 100◦ C (Campbell and Jarvis, 1984), 150–250◦ C
(Herzberg and Asimow, 2008) or up to 400◦ C (Sleep and Windley, 1982). Long-term cooling rates are estimated to be in the range of 45◦ C Ga−1 (Labrosse and Jaupart, 2007) and
70◦ C Ga−1 (Abbott et al., 1994). However, the secular cooling rate of the mantle does
not have to be constant (Korenaga, 2008) and other processes may influence the thermal
structure of the mantle on shorter timescales.
Regional variations in mantle potential temperature do not only affect the chemical composition of MORB (Klein and Langmuir, 1987) but also result in regional variations in
the subsidence rate of the seafloor (Hayes, 1988) and axial depth of the mid-ocean ridge
(e.g., Marty and Cazenave, 1989; Calcagno and Cazenave, 1994). Based on the geochemistry of ancient MORB, Humler et al. (1999) proposed a globally higher mantle potential
temperature of more than 50◦ C for the time period before 80 Ma. Possible explanations
for this global temperature difference include a ‘mantle-avalanche’ (Machetel and Humler,
2003) or the on-going breakup of the supercontinent Pangaea (mantle temperature as a
function of distance between continent and mid-ocean ridge; Humler and Besse, 2002).
Based on my new data, changes in mantle potential temperature during the Mesozoic are
not a global phenomenon but are restricted to the Atlantic and possibly Indian oceans as
a result of continental insulation (see chapter 3).
1.2 Geochemistry of oceanic basalts
In this section, I will discuss the processes that control the chemical composition of oceanic
basalts apart from changes in mantle potential temperature. As mentioned above, the
Earth’s mantle is highly heterogeneous and the composition of the mantle source thus
has also a major influence on the chemical composition of the melt erupted. However, numerous processes are affecting the primary melt prior to eruption, obscuring the original
signature from the mantle source. These include melt aggregation and mixing, contamination by melt-rock interaction and fractional crystallisation (e.g., O’Hara, 1965; DePaolo,
1981; Klein and Langmuir, 1987). In order to constrain mantle source compositions, melting processes, or mantle evolution it is thus essential to correct for these processes if
possible, or if not, to carefully discuss potential bias effects.
1.2.1 Melting in the mantle
The large lithostatic pressure from the overlying oceanic or continental lithosphere normally prevents the mantle from melting. However, there are three principle ways how to
melt the mantle: increasing the mantle potential temperature, adding volatiles that lower
Dissertation P.A. Brandl
5
1. Introduction
the mantle solidus temperature or adiabatic upwelling as a result of plate separation and
decreased lithostatic pressure.
Figure 1.4: Anhydrous P-T diagram for mantle lherzolite KLB1 showing the stabilitiy fields of garnet, spinel and plagioclase and mantle solidus and liquidus. Adiabatic upwelling (example adiabat shown in
green) causes decompressional melting.
The degree of partial melting (F) is
thereby controlled by the depth at
which the adiabat intersects the mante
solidus and thus mantle potential temperature (Tp ). Figure modified from
Thompson and Gibson (2000).
The ‘classical’ view on melting at mid-ocean ridges is that of passive upwelling of a peridotite mantle (e.g., O’Hara, 1965; Cann, 1968; Moore, 1970; Bottinga and Allègre, 1973;
Wyllie, 1973). The separation of plates at the ridge axis induces adiabatic upwelling in
the underlying mantle that causes partial melting (Fig. 1.4). The mantle solidus will be
intersected at different depths, depending on initial mantle potential temperature, but
generally at about 80 to 90 km depth, close to the garnet-spinel transition zone. The
Mantle Electromagnetic and Tomography (MELT) experiment traced basaltic melt over
a several hundred kilometres broad region and to depths greater than 100 km underneath
the East Pacific Rise (The MELT Seismic Team, 1998). Melting in the presence of water
can start even deeper (150–200 km) and the shape of the melting region is controlled
by spreading rate and (a-)symmetry of spreading (The MELT Seismic Team, 1998). The
melting rate increases towards the ridge axis and the overall degree of partial melting is
commonly in the order of 8–20% (Klein and Langmuir, 1987). When melt is retained in
the ambient upwelling mantle, this will lead to a higher buoyancy, inducing an additional
component of active (or dynamic) upwelling to the normal passive upwelling underneath
mid-ocean ridges (e.g., Macdonald et al., 1988; Langmuir et al., 1992; Lundstrom et al.,
1998).
6
Dissertation P.A. Brandl
1. Introduction
In contrast to mid-ocean ridges, melting in mantle plumes and hotspots is controlled
by active mantle upwelling. (Deep) Mantle plumes probably initiate through instabilities
in the D”-layer, a mantle layer close to the core-mantle boundary (e.g., Morgan, 1972;
Hofmann and White, 1982; DePaolo and Manga, 2003 and references therein). The discussion whether deep mantle plumes are the only potential origin for hotspots or whether
some of these may originate at the mantle’s 670 km discontinuity (see discussion in DePaolo and Manga, 2003) is a matter of debate. However, overall differences in density and
viscosity allow the hot plume material to ascent through the ambient mantle. During this
process mantle plumes get their ‘classical’ plume shape with a broad and very hot plume
head and a narrow plume tail as demonstrated by analogue models (Campbell and Griffiths, 1990). Melting in (hot) mantle plumes starts significantly deeper than at mid-ocean
ridges, generally within the garnet stability field (>80 km) due to the intersection of the
mantle solidus at greater temperatures and thus at grater depth.
Moreover, the impact of a mantle plume head to the lithosphere can result in sublithospheric melting and widespread volcanism. Many flood basalt provinces, such as the Deccan Traps, the Siberian Traps or the North Atlantic Volcanic Province, are probably
related to the initial arrival of a mantle plume head (e.g., Morgan, 1972; Courtillot et al.,
1986; Cox, 1989; White and McKenzie, 1989, 1995; Campbell and Griffiths, 1990). In
contrast to the plume head, the plume’s tail will supply locally focused but long-living
magmatism. Volcanic chains or seamount trails, such as the Hawaiian-Emperor or the
Louisville Seamount chains, are formed where the overriding lithosphere plate is moving
above such a stationary plume tail. Another classical example for hotspot volcanism with
a seismically imaged, deep-rooted mantle plume is Iceland (e.g., Tryggvason et al., 1983;
Wolfe et al., 1997; Allen et al., 2002).
More recently, volatiles (especially H2 O and/or CO2 ) are considered to play an important
role for melting the mantle (e.g., Wyllie, 1971; Kushiro, 1972; Wyllie, 1977; Yaxley and
Brey, 2004; Hammouda, 2003; Dasgupta et al., 2004; Dasgupta and Hirschmann, 2006).
One example where the volatiles may influence or even induce melting is the Azores
archipelago. It is still a matter of active debate whether magmatism in the Azores is
resulting from melting by active upwelling of hot mantle material or by the presence
of substantial amounts of volatiles in the mantle (‘hotspot versus wetspot’; e.g., Asimow
et al., 2004; Asimow and Langmuir, 2003). Geochemical studies have shown that water (or
volatiles in general) have to be present in the Azores mantle source since mantle potential
temperatures are too low to explain the observed degrees of partial melting (Beier et al.,
2012a). A recent study on lavas from Santa Maria has shown that substantial amounts of
CO2 must be present in the Azores mantle source and evidence for melting of carbonated
peridotite is preserved through the extremely low degrees of partial melting (Beier et al.,
Dissertation P.A. Brandl
7
1. Introduction
2012b).
1.2.2 Melt extraction and magma mixing
The final depth of melting is another important factor that controls the melting process
and the chemical composition of the lavas erupted (Shen and Forsyth, 1995). The melting
column does not necessarily extent to the lower boundary of the lithosphere but could
cease at even greater depth. Some possible explanations for this observation include fast
melt extraction, e.g., through olivine channels (e.g., Suhr, 1999; Suhr et al., 2003; Kelemen et al., 2000), melt extraction and chemically isolated transport when a critical volume
is reached (see discussion below) or conductive cooling from the surface combined with
high-pressure crystallisation (especially at slow-spreading ridges and large-offset fracture
zones; e.g., Shen and Forsyth, 1995; Michael and Cornell, 1998; Eason and Sinton, 2006).
Initial and final depths of melting were first studied in detail by the pioneering work
of Shen and Forsyth (1995). They investigated the relationship between fractionationcorrected major element data, major and trace element ratios sensitive to degree of melting and source-enrichment and geophysical parameters such as spreading rate or crustal
thickness. This would have important implications for the range of global variation in
mantle potential temperature as recorded in the geochemistry of MORB.
The process of magma mixing during transport and storage and the melt transport are
other important aspects that influence the chemical composition of erupted basalts (Fig.
1.5. These processes can be studied in more detail by tracing and modelling melting in a
chemically and lithologically heterogeneous mantle (e.g., composed of pyroxenite and peridotite). The major question for the melt extraction processes is whether enriched (and
deep) melts equilibrate with the surrounding matric during ascent or not (equilibrium
versus disequilibrium melting). It has been suggested that melts formed from pyroxenite
veins larger than 0.1 to 1 m width can escape their peridotitic source (Kogiso et al., 2004)
but melts from smaller pyroxenite bodies are either not resolvable (Stolper and Asimow,
2007) or diffusively equilibrated with the peridotite matrix (Stracke and Bourdon, 2009).
Nevertheless, large-scale melt-rock interaction as suggested by Sobolev et al. (2005) can
“. . . only occur in a regime that permits large-scale reactive porous flow.” (Stracke and
Bourdon, 2009).
Several studies (e.g., Kelemen, 1990; Hart, 1993; Kelemen et al., 1997; Spiegelman and
Kelemen, 2003) including uranium series studies (e.g., Bourdon et al., 1996; Sims et al.,
1999, 2002; Lundstrom, 2000; Rubin et al., 2005, 2009; Stracke et al., 2006) of oceanic
basalts have shown that melt transport occurs at least not solely by porous flow but also
8
Dissertation P.A. Brandl
1. Introduction
Figure 1.5: Schematic model of the magmatic plumbing system underneath midocean ridges. Enriched lithologies start melting
deeper (than ambient depleted mantle) forming enriched melts. With progressive melting these enriched melts mix with more depleted melts formed at
the edges of the melt channels. These melts mix continuously during transport and during (fractional)
crystallisation in magma chambers. Figure modified
from Maclennan (2008).
includes fast melt transport through melt channels (channelised flow melting; Fig. 1.5).
A two-porosity melting scenario with porous, reactive melt flow and disequilibrium melt
extraction through high-porosity channels when the overall degree of partial melting is
increasing (and thus the overall degree of mantle depletion) may be also plausible (e.g.,
Kelemen et al., 1997; Lundstrom, 2000; Jull et al., 2002; Sims et al., 2002; Rubin et al.,
2009; Stracke and Bourdon, 2009).
Magma mixing also plays an important role during the formation of MORB. Melt inclusion studies have shown that the erupted lavas always represent mixtures of chemically
very heterogeneous melts. For example, melt inclusions from one single hand-specimen
from the Reykjanes Peninsula in southwest Iceland recorded 50–90% of the variation in
208
Pb/206 Pb observed in MORB from the entire Mid-Atlantic Ridge (Maclennan, 2008).
It is remarkable that major and trace element and radiogenic isotope data show the same
trends in melt inclusions and whole-rock samples, with melt inclusions extending the
range of heterogeneity towards more depleted compositions (e.g., Sobolev and Shimizu,
1993; Shimizu, 1998; Maclennan, 2008). Thus, melt inclusions give evidence that the
mantle source, even of MORB, is extremely heterogeneous but with progressive melting
and transport, the ascending melt becomes progressively homogenised (Fig. 1.5). Rubin
Dissertation P.A. Brandl
9
1. Introduction
et al. (2009) showed that this effect of homogenisation and differentiation is also a function of spreading rate and the rate of magma supply, respectively. Fast-spreading ridges
such as the East Pacific Rise generally have a higher magma supply rate compared to
slow-spreading ridges, resulting in more differentiated but also more homogeneous lava
compositions.
As a result, global compilations of mid-ocean ridge basalts show chemical correlations
that are regional and mainly of compositional origin and correlations that are mainly
controlled by variations in mantle potential temperature (e.g., Klein and Langmuir, 1987;
Langmuir et al., 1992; Niu and O’Hara, 2008). To what extent chemical variations in
MORBs reflect source heterogeneity or mantle temperature is still a matter of active debate, but there is overall consensus that at least some of this variability is the result of
changes in mantle temperatures.
1.2.3 Shallow-level processes
The composition of oceanic basalts is strongest affected by shallow level processes, such
as differentiation by fractional crystallisation in magma chambers or by wallrock assimilation (e.g., Bowen, 1928; Allègre and Minster, 1978; Taylor Jr., 1980; DePaolo, 1981;
Albarède, 1995) that need to be considered before investigating the ‘deeper’ processes.
The majority of MORB is composed of subalkalic tholeiitic lavas (e.g., Engel and Engel, 1964a,b; Engel et al., 1965), but even these primitive rocks do not represent the
composition of a primary mantle melt (e.g., O’Hara, 1968; Falloon and Green, 1988).
During magma aggregation and storage, the melt cools and as a consequence mineral
crystals fractionate from the liquid in a complex kinetic process. The most common mineral assemblage in MORB is olivine and plagioclase ± clinopyroxene ± opaque minerals
(e.g., Green and Ringwood, 1967; Miyashiro et al., 1969; Shido et al., 1971). Starting with
a primary melt, olivine is usually the first fractionating phase from the liquid (about 10
to 25% at 10 kbar → ‘olivine control line’; e.g., Falloon and Green, 1987; Putirka et al.,
2007). The proportions of the three phases olivine, clinopyroxene and plagioclase (with
special emphasis on the plagioclase-clinopyroxene ratio) is dependent on magma composition and pressure of crystallisation (e.g., O’Hara, 1968; Michael and Cornell, 1998;
Herzberg, 2004). A recent study by O’Neill and Jenner (2013) proposed that much of the
chemical variation (especially of Rare Earth Elements, REE) in MORB could be explained
by variations in the ratio of plagioclase to clinopyroxene as a consequence of cycles in the
ocean ridge magma chamber (magma replenishment rate) and based on the fact that REE
are slightly less incompatible in clinopyroxene compared to plagioclase. Several methods
to correct the geochemistry of MORB for the effects of fractional crystallisation have been
10
Dissertation P.A. Brandl
1. Introduction
proposed (e.g., Klein and Langmuir, 1987; Langmuir et al., 1992; Niu et al., 1999; Kelley
et al., 2006; Herzberg et al., 2007; Herzberg and Asimow, 2008; Niu and O’Hara, 2008).
Probably one of the most precise methods to account for these effects requires the study
of each distinct batch of samples and their individual Liquid Line of Descent (LLD). It
would then be possible to track back the chemical composition along the LLD to a point
that represents a primitive magma composition. In many cases this method cannot be
applied since many MORB sample suites simply do not record a wide range of chemical
differentiation to identify the corresponding slope and shape of the LLD suitably. Thus,
a different approach to correct for the chemical effects of fractional crystallisation is required for MORB.
Olivines are probably the most common xenocryst phase in basalts and allow to constrain the primary melt composition by their systematic exchange of Fe-Mg. Primary
mantle olivines have forsterite contents of about 89–92 (e.g., Niu and O’Hara, 2008; Kelley et al., 2006; Perfit et al., 1996). Assuming a partition coefficient Kol–liq
D(Fe-Mg) = 0.30±0.03
(Roeder and Emslie, 1970) would result in a primitive melt (glass) composition of Mg#
= 72 (Mg# = 100 x Mg2+ /(Mg2+ + Fe2+ )). As a result, geochemical data of MORB have
to be corrected to a common concentration in MgO to compare different sets of samples
at a reliable degree of fractionation (e.g., 8.0 wt. % MgO; Klein and Langmuir, 1987) but
do not necessarily represent a primitive melt composition. Primitive melt composition of
MORB samples can best be recalculated following either the polynomial approach of Niu
et al. (1999) and Niu and O’Hara (2008) or following the three-step linear-incremental
method outlined by Kelley et al. (2006).
Assimilation of crustal or lithospheric material is another important process that influences the chemical composition of magmatic rocks. These reactive melt-rock interactions
can potentially enhance the compositional variability with the result that the erupted
lavas do not reflect their primary source composition (Rubin et al., 2009). Melt-rock interaction can re-fertilise a mantle that had been previously depleted by melt extraction
in the vicinity of mid-ocean ridges (e.g., Kelemen et al., 1992; Niu and O’Hara, 2003) or
by melts of very low degrees of partial melting in the subcontinental lithosphere (e.g.,
O’Reilly and Griffin, 1988; Bodinier et al., 1996). To what extent melt-rock reactions influence the composition of MORB is still a matter of active debate. In contrast to these
chemical reactions, the assimilation of hydrothermally altered crust into the magma can
be traced using the Cl/K ratios (Michael and Cornell, 1998) or oxygen isotope ratios (e.g.,
Staudigel et al., 1995, 1981).
Assimilation and Fractional Crystallisation (AFC) are important processes affecting the
geochemical composition of erupted lavas. Numerous studies have tried to determine the
Dissertation P.A. Brandl
11
1. Introduction
chemical influence of AFC processes on MORB and published methods how to correct
for these effects (e.g., DePaolo, 1981; Klein and Langmuir, 1987; Niu et al., 1999; Kelley
et al., 2006). In this study, I applied these methods to drilled samples of oceanic crust for
which it is additionally important to constrain the eruptive and formation processes of
the oceanic crust itself.
1.3 Structure of the oceanic igneous crust
The structure of the oceanic crust is considered as a relatively simple layered sequence of
rock types that reflect the accretionary process at the mid-ocean ridges (Fig. 1.6; Karson,
2002). Based on field studies on ophiolites (Boudier and Nicolas, 1985), the oceanic crust
is commonly subdivided into three layers comprising the sedimentary layer 1, the volcanic
layer 2 and the plutonic layer 3 (Fig. 1.6; Cann, 1974). The sediment thickness of layer
1 is depending on age of the crust and sedimentation rate. The total thickness of the
igneous oceanic crust (extrusive layer 2A, sheeted dyke complex/intrusive layer 2B and
gabbroic layer 3) is about 7.1±0.8 km (White et al., 1992) and is relatively constant for
full spreading rates higher than about 15 mm a−1 (Bown and White, 1994).
Figure 1.6: Simplified
crosssection of the oceanic crust. a)
internal structure of the oceanic crust
and b) its interpretation into layers:
sediments (layer 1), basaltic pillow
lavas (layer 2A), sheeted dyke complex
(layer 2B), gabbroic rocks (layer 3).
c) Shows exemplary outcrop fotos of
pillow lavas from Macquarie Island
(top), sheeted dyke complex from
the Semail ophiolite (middle) and
gabbroic rocks from the Bay of Island
ophiolite (bottom). Figure modified
from Karson (2002).
Mid-ocean ridge basalt is the most common rock type in the volcanic section of the igneous crust and the equivalent plutonic rocks (gabbro) make up the most volume of the
12
Dissertation P.A. Brandl
1. Introduction
intrusive layer 3 (Mutter and Mutter, 1993). These rocks form from relatively high degrees
of partial melting (∼8–20%) over a pressure range of about 5–16 kbar (Klein and Langmuir, 1987). For spreading rates higher than 15 mm a−1 , mantle potential temperature is
directly controlling the physical conditions of mid-ocean ridges, such as the thickness of
the oceanic crust produced and the waterdepth of the ridge (e.g., Klein and Langmuir,
1987; McKenzie and Bickle, 1988; Langmuir et al., 1992; Bown and White, 1994; Su et al.,
1994). The mantle potential temperature also controls the degree of partial melting as a
function of depth (pressure) at which the adiabat intersects the mantle solidus (Fig. 1.4;
e.g., Green and Liebermann, 1976; Forsyth, 1977). Thus, the geochemical composition of
MORB (that is essentially controlled by the degree of partial melting) provides a record
of mantle potential temperatures.
Figure 1.7: Morphology of fast- and slow-spreading mid-ocean ridges. Ridge morphology of a)
a fast-spreading ridge (East Pacific Rise at 3◦ S) and b) a slow-spreading ridge (Mid-Atlantic Ridge at
37◦ N). Note that slow-spreading ridges are signifcantly influenced by tectonic processes forming a wide
axial rift valley. Figure modified from Standish and Sims (2010).
As mentioned above, the total thickness of the oceanic crust is relatively independent
from spreading rate for spreading rates higher than ∼15 mm a−1 . However, spreading rate
in combination with magma supply rate influences both the morphology of the mid-ocean
ridge (axial high versus rift valley, width of the neovolcanic zone; Fig. 1.7) and thus also
the structure of the oceanic crust (e.g., Macdonald, 1982; Phipps Morgan and Chen, 1993;
Small, 1998). The spreading rate controls the depth from the top of the igneous basement
to the axial low-velocity zone (representing the magma chamber and, after crystallisation, gabbroic layer 3) and thus the thickness of volcanic layer 2 (e.g., Purdy et al., 1992;
Dissertation P.A. Brandl
13
1. Introduction
Phipps Morgan and Chen, 1993; Carbotte et al., 1998). The difference in thickness of
volcanic layer 2 may best be explained by differences in the accretion processes between
slow- and fast-spreading ridges. Slow-spreading ridges, such as the Mid-Atlantic Ridge,
are characterised by axial valleys and wide neovolcanic zones (Fig. 1.7b; Macdonald, 1982
and references therein). Axial valleys may efficiently trap all erupted lavas (Canales et al.,
2005) and, in combination with the wide neovolcanic zone and slow-spreading rates (resulting in a long residual time of newly formed oceanic crust within this zone), may
account for long-lasting, periodical dyking and eruption events that build up the volcanic
section. As a result, on-axis thickness of layer 2A correlates positively with depth to the
axial magma chamber (e.g., Buck et al., 1997; Canales et al., 2005). In contrast, fastspreading ridges often form axial highs (Fig. 1.7), representing axial summit volcanoes
(calderas), and the lava flows are able to escape the neovolcanic zone by flowing off-axis
over long distances. It is thus not surprising that off-axis lava flows may be responsible
for the off-axial thickening of the uppermost oceanic crust at fast-spreading ridges, such
as the East Pacific Rise (e.g., Hooft et al., 1996; Carbotte et al., 1998; Canales et al., 2005).
Summarising, the structure and chemical composition of the oceanic crust are affected by
numerous physical and chemical parameters, such as source composition, mantle temperature, spreading rate and fractional crystallisation. However, since mid-ocean ridge basalts
are the most common rock type of oceanic crust accessible from the surface, geochemical
studies of MORBs allow to infer on these processes and to determine spatial and temporal
variations. New oceanic crust is formed continuously at the ridge axis and transported
away from the axis through the spreading process. With increasing age, sediments bury
the igneous oceanic crust and ocean drilling provides the only opportunity to access old
igneous oceanic crust.
In this study, I will present some of my results on mid-ocean ridge processes, mantle
heterogeneity and thermal evolution of the Earth’s mantle that have been obtained by
studying the geochemistry of oceanic basalts.
14
Dissertation P.A. Brandl
2. Aims of the study
2 Aims of the study
2.1 Constraining the thermal evolution of the upper
mantle
Major changes in the length or activity (either by spreading-rate or magma production
rate) of the global mid-ocean ridge system have a significant influence on eustatic sealevel
(Gaffin, 1987), seawater chemistry (Hardie, 1996), the climate (Berner and Kothavala,
2001) and in final consequence the biosphere (Larson, 1991a) over long timescales (10–
100 Ma). Geophysical studies allow to infer on temporal variations in the total ridge
length and spreading rate, e.g., through paleomagnetic data and plate reconstructions
(e.g., Müller et al., 2008b,a; Seton et al., 2009; Cogné et al., 2006), but the magmatic
production rate is recorded in the geochemical composition of MORB and the thickness
of the oceanic crust only (e.g., Klein and Langmuir, 1987; Langmuir et al., 1992; White
et al., 1992; Bown and White, 1994; Humler et al., 1999).
Information on mantle melting conditions in the past are thus provided by sampling
and studying ancient MORB, e.g., by drilling into old oceanic igneous crust. An initial
compilation of geochemical data of 17 DSDP and ODP sites drilled into the volcanic
section of old oceanic crust was interpreted to reflect a globally higher mantle potential
temperature of more than 50◦ C in the time prior to about 80 Ma (Humler et al., 1999).
The inferred difference in mantle potential temperature is greater than what would be
the result of normal secular cooling of the mantle (45–70◦ C Ga−1 ; Abbott et al., 1994;
Labrosse and Jaupart, 2007).
However, Humler et al. (1999) owe an explanation for the large difference in mantle potential temperature recorded in their study. Later studies argued for a Cretaceous ‘mantle
avalanche’ (Machetel and Humler, 2003) or heat transfer from the subcontinental mantle
to the mid-ocean ridges (Humler and Besse, 2002). The aim of my study was to infer the
thermal evolution of the upper mantle in more detail. The study focused on fresh glasses of
old seafloor since fresh glasses reflect the chemical composition of the melt more precisely
than whole-rock data that may be compromised by variable amounts of phenocrysts and
alteration.
Dissertation P.A. Brandl
15
2. Aims of the study
2.2 Insights into mantle composition and melting
The conversion of MORB chemistry to mantle potential temperature is complicated by the
fact that MORBs are chemically heterogeneous even in the absence of nearby hotspots
(e.g., Allègre and Turcotte, 1986; Arevalo Jr. and McDonough, 2010; Donnelly et al.,
2004; Hirschmann and Stolper, 1996; Niu et al., 1999; Phipps Morgan and Morgan, 1999;
Schilling et al., 1983; Zindler and Hart, 1986. The range in MORB chemistry results from
melting a heterogeneous mantle. Enriched mantle material contains significant concentrations of highly incompatible elements and is likely dispersed as ‘streaks’ or ‘veins’ in a
volumetrically dominant matrix of depleted ‘ambient’ mantle material (e.g., Batiza and
Vanko, 1984; Sleep, 1984; Zindler et al., 1984; Prinzhofer et al., 1989). Thus, even small
proportions of enriched mantle in the ambient depleted mantle may have a significant
influence on the geochemical composition of MORB.
It is thus essential for the understanding of MORB composition and mantle geochemistry to precisely constrain the melting processes (such as melt productivity and mixing)
and source composition. Several studies argue for lower solidus temperatures and higher
melt productivity of relatively fertile mantle material (enriched in incompatible elements)
compared to the depleted matrix (e.g., Ito and Mahoney, 2005a,b; Meibom and Anderson, 2004; Sleep, 1984; Prinzhofer et al., 1989; Salters and Dick, 2002). However, MORBs
are generally characterised by overall high degrees of partial melting of large volumes of
the underlying mantle combined with effective magma mixing and homogenisation during
magma ascent and storage in large magma chambers. As a result, lavas erupted at active
mid-ocean ridges are unlikely to preserve the full range of geochemical variability present
in the upper mantle (e.g., Bryan, 1983; Dungan and Rhodes, 1978; Rhodes et al., 1979;
Rubin and Sinton, 2007; Sinton and Detrick, 1992; Walker et al., 1979).
I thus focused on the geochemistry of a near-ridge seamount (Seamount 6; chapter 4)
and a dying spreading axis (Galapagos Rise; chapter 5) because both of these settings are
characterised by smaller volumes and degrees of partial melting on average. These lavas
provide thus a higher potential to preserve the initial chemical variability of the mantle.
The main aim of my study at Seamount 6 (chapter 4) was to record the full chemical
variability of a distinct location that may lead to a precise ‘snapshot’ of the compositional heterogeneity in the underlying mantle. In contrast, the geochemical study of postspreading volcanic rocks (chapter 5) provide the unique opportunity to infer on the effects
of source heterogeneity and melting processes underneath spreading centres depending on
the rate of spreading.
16
Dissertation P.A. Brandl
2. Aims of the study
2.3 Studying volcanic processes at mid-ocean ridges
To improve our understanding of the magmatic processes that form the oceanic crust
it is essential to study the geochemistry of distinct eruptive units at mid-ocean ridges
in order to precisely constrain the full geochemical composition recorded in the oceanic
crust. Single eruptive events have so far mainly been studied at the East Pacific Rise, the
Juan de Fuca Ridge and Iceland and indicate that geochemical variations are observed
not only on an inter-flow but also on an intra-flow scale (e.g., Hall and Sinton, 1996;
Perfit and Chadwick, 1998; Sigmarsson et al., 1991; Sinton et al., 2002; Maclennan et al.,
2003). In contrast, only few studies focused on (or even observed) eruptive processes at
slow-spreading ridges in terms of their chemical variation (e.g., Sinton et al., 2002; Stakes
et al., 1984) or evolution over several thousand years (e.g., Rubin and Macdougall, 1990;
Sturm et al., 2000).
Nevertheless, geochemical studies of individual eruptions may help to resolve the general
structure of the magmatic plumbing system at individual ridge segments. An interconnection of ‘distinct’ magma reservoirs is possible (Magde et al., 2000), but melt transport
may also be focused to the segment centres (e.g., Macdonald et al., 1988; Abelson et al.,
2001; Magde et al., 2000) or, at least at high magma supply rates, dispersed over the
entire length of the segment (Tucholke et al., 1997).
In chapter 6, I will present a combined study of geological observations, petrology and
geochemistry of a young volcanic field on the southern Mid-Atlantic Ridge. This study
aims to contribute to our general understanding of the magmatic plumbing system and
eruptive processes at slow-spreading ridge in order to infer the structure and geochemistry
of the oceanic crust as a whole.
Dissertation P.A. Brandl
17
3. High mantle temperatures following rifting
3 High mantle temperatures following
rifting caused by continental
insulation
Philipp A. Brandl, Marcel Regelous, Christoph Beier and Karsten M. Haase
GeoZentrum Nordbayern, Friedrich-Alexander-Universität Erlangen-Nürnberg, Schlossgarten 5, 91054 Erlangen, Germany
3.1 Abstract
On geological time-scales, the distribution of continents is thought to influence mantle
temperature, and thus global magmatism (Anderson, 1982; Grigné and Labrosse, 2001;
Lenardic et al., 2005; Zhong and Gurnis, 1993). It has been proposed that continental insulation effects may be responsible for periods of significantly increased magmatism during
continental breakup, including flood basalt volcanism and associated environmental and
biologic effects (Coltice et al., 2007, 2009). Although numerical models and laboratory
studies predict temperature effects of several tens of degrees, direct geological evidence in
support of mantle warming due to continental insulation has been lacking. Here we present
electron microprobe major element analyses of ancient mid-ocean ridge basalt glasses from
drillsites in Mesozoic-Cenozoic ocean crust in the Atlantic and Pacific, which preserve a
record of upper mantle temperature and melting processes over the past 170 Ma. We
show that temporal chemical changes in the lavas erupted at spreading ridges following
continental rifting and breakup can be used to quantify the continental insulation effect.
We find that in the Atlantic, the upper mantle temperature immediately after rifting was
up to about 150◦ C higher than present, and a thermal anomaly persisted for 60-70 Ma.
Higher mantle temperatures beneath young ocean basins would result in higher global
sealevel and enhanced element fluxes at mid-ocean ridges.
Dissertation P.A. Brandl
19
3. High mantle temperatures following rifting
3.2 Introduction
Approximately 75% (about 21 km3 a−1 ) of magmatism on Earth occurs along the
60,000 km long mid-ocean ridge (MOR) system. Formation of the oceanic crust at MOR
spreading centres and its subsequent evolution has an important influence on sealevel
(Gaffin, 1987), the carbon cycle (Berner and Kothavala, 2001) and seawater chemistry
(Hardie, 1996) on timescales of 10–100 Ma. Magmatic activity at spreading ridges is responsible for about 30% of the annual volcanic CO2 output (Marty and Tolstikhin, 1998),
and elemental transfer between seawater and lithosphere due to hydrothermal processes
play an important role in the global geochemical cycles of many elements (Elderfield and
Schultz, 1996). Changes in the total length of the MOR system and spreading rate, or
the temperature of the underlying mantle and the mantle melting process over the lifetime of ocean basins could therefore have important effects on eustatic sealevel, seawater
composition, global climate and the biosphere (Larson, 1991b).
Whereas temporal variations in the total ridge length and spreading rate can be determined from paleomagnetic data and plate reconstructions, past changes in the conditions
of mantle melting beneath ridges can be identified through chemical analysis of ancient
seafloor lavas. Humler et al. (1999) compiled data for samples from 17 DSDP and ODP
sites which drilled volcanic basement in the Atlantic and Pacific Oceans. After correcting
for the chemical effects of fractional crystallisation, they found apparent differences in the
major element compositions of ancient (>80 Ma) mid-ocean ridge basalts (MORB), and
those from active spreading centres. For a given MgO content, MORB older than about
80 Ma have higher FeO and lower Na2 O. These chemical differences were interpreted to
result from higher mantle temperatures in the Mesozoic, resulting in higher degrees of
melting at higher average pressure (Humler et al., 1999; Machetel and Humler, 2003).
The inferred temperature difference (about 50◦ C) is too great to represent normal secular
cooling of the mantle. The origin and significance of the temperature difference is therefore
unclear, especially because the dataset includes altered and phenocryst-rich whole-rock
samples, which cannot be used to infer melt compositions reliably.
3.3 Methods
The major element compositions of fresh volcanic glasses from DSDP/ODP/IODP drillcores were measured using a JEOL JXA-8200 Superprobe electron microprobe at the
GeoZentrum Nordbayern. For comparison, major element data for 9,800 samples of MORB
glass from active spreading ridges was compiled from the Smithsonian Abyssal Volcanic
Glass Data File (AVGDF) and PetDB. Both datasets were screened for the effects of
alteration, inter-laboratory bias, and duplicate analyses (see section 3.6). In order to in-
20
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
vestigate mantle melting and source effects on MORB chemistry, we corrected the data for
the effects of low-pressure fractional crystallisation. After excluding from consideration all
samples having MgO values less than 7.0 wt. % (see section 3.6), each individual sample
was corrected for the effects of fractional crystallisation to an MgO of 8.0 wt. % (Klein
and Langmuir, 1987). Primary magma compositions (Kelley et al., 2006) were used to
estimate temperatures using established major element geothermometers (Kelley et al.,
2006; Herzberg et al., 2007). Full details are available in section 3.6.
3.4 Results & Discussion
We carried out a systematic geochemical study of samples of ancient MORB from
30 drillsites in the Atlantic and Pacific which were drilled into volcanic basement older
than 6 Ma. We analysed exclusively fresh volcanic glasses in order to avoid the effects
of seawater alteration and crystal accumulation. Our new dataset includes 340 samples
(183 from the Atlantic, 157 from the Pacific, see section 3.6), which range in age up to
170 Ma (Pacific) and 166 Ma (Atlantic). For comparison, we compiled a dataset for zeroage MORB glasses from active spreading ridges in the Atlantic and Pacific, using data
from the Smithsonian Institution and PetDB (see section 3.6).
Fractionation-corrected Na2 O (Na8 ), Ti8 , Fe8 and CaO/Al2 O3 values (Klein and Langmuir, 1987) for the drilled samples generally lie within the fields defined by zero-age
MORB from the respective ocean basins, but are displaced to the lower Na8 end of these
fields (Fig. 3.1). Samples with the lowest Na8 values have the lowest Ti8 and the highest
Fe8 and CaO/Al2 O3 (Fig. 3.1). Atlantic drillsites record an increase in average Na8 from
values of about 1.7 wt. % at 165 Ma, to values typical of active Atlantic spreading centres
(2.5 wt. %) at sites younger than 100 Ma (Fig. 3.2). Average Fe8 values decreased from
around 11.0 wt. % to about 9.5 wt. % over the same time period. In contrast, samples
from Pacific drillsites show no systematic change in composition with time over the past
170 Ma (Fig. 3.2), although drill samples have on average lower Na8 and higher Fe8 than
the mean for young Pacific MORB.
Apparent differences in the average fractionation-corrected major element compositions of
young and ancient MORB could arise from sampling bias. The drillsites we have sampled
contain a record of magmatism at a single site, likely over a period of <50 ka. Studies
of active spreading ridges show that MORB chemistry varies significantly on the scale of
individual ridge segments (Niu et al., 2001), and over periods of a few tens of ka at a single
location (Regelous et al., 1999). Vast lengths of the active MOR system are only sparsely
sampled, and approximately 80% of all Pacific samples in our zero-age dataset are from
Dissertation P.A. Brandl
21
3. High mantle temperatures following rifting
Figure 3.1: Fractionation-corrected major element compositions of lavas drilled from ancient oceanic crust. Major element compositions of ancient MORB glasses (coloured symbols) and
zero-age MORB from active spreading centres (grey symbols) from the Atlantic and Pacific, corrected
for the effects of fractional crystallisation to 8 wt. % MgO. Diamond symbols represent new data from
this study, triangles previously published data. Depth average compositions (Table A5, Appendix) for
active spreading ridges (hexagons) shown for comparison. The oldest oceanic crust in the Atlantic is
characterised by low Na8 , Ti8 , high Fe8 and CaO/Al2 O3 compared to most zero-age Atlantic MORB,
whereas the oldest drilled lavas in the Pacific have similar compositions to zero-age Pacific MORB.
the northern East Pacific Rise and the Galapagos Rise Spreading Centre, a combined
ridge length of about 4,000 km. For these reasons, a rigorous statistical comparison of the
two datasets is not possible. Nevertheless, assuming an average depth for the global MOR
system of 2,600 m (Stein and Stein, 1992) the corresponding Na8 value is 2.48 wt. % (Niu
and O’Hara, 2008), and the average Na8 for Pacific and Atlantic MORB in our zero-age
dataset is similar (2.55 wt. %). Thus, the fact that zero-age MORB with Na8 values as
low as 1.70 wt. % are rare even though atypically shallow regions of the ridge system with
low Na8 (e.g., Reykjanes Ridge) may be over-represented in the zero-age MORB compilation, suggests that the chemical differences between the two datasets are significant.
Lavas erupted at the active ridge axis will eventually make up the lowermost parts of the
mature ocean crust. In contrast, drillsites in ancient ocean crust preferentially sample the
uppermost, youngest lavas emplaced at that location (most cores we have sampled were
drilled into only the uppermost 200 m of basement), and may include larger-volume flows
that flow down the ridge flanks, or lavas erupted off-axis. If significant chemical differences exist between flows emplaced on- and off-axis, a meaningful comparison of drilled
and dredged samples may not be possible. However, there is no convincing evidence from
detailed geochronological and geological studies for any significant compositional difference between lavas dredged from the ridge axis and those exposed on the ridge flanks
22
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
Figure 3.2: Temporal variations
in composition of MORB from the
Atlantic and Pacific. Histograms at
left show Na8 distribution for zero-age
MORB from (a) the Atlantic and (b)
the Pacific Ocean. Ancient Atlantic
MORB (coloured symbols) record an
increase in Na8 with decreasing age,
from values as low as 1.6 wt. % in
150–160 Ma MORB, which are rare in
zero-age samples, to values of around
2.7 wt. %. In contrast, there has been
no systematic change with time in
the compositions of Pacific MORB.
Symbols as in Figure 3.1.
(Waters et al., 2011), and eruption of chemically-distinct lavas on the ridge flanks would
result in a systematic difference between MORB from active ridge axes and those from all
drillsites, rather than the temporal change over 170 Ma observed in the Atlantic samples.
Variations in the major element composition of MORB over long timescales could result either from changes in the composition of the mantle source (Janney and Castillo,
1997, 2001), or from changes in the physical conditions of melting (Humler et al., 1999;
Machetel and Humler, 2003; Fisk and Kelley, 2002). If the low Na8 of ancient Atlantic
samples resulted from melting of buoyant, refractory mantle (Niu and O’Hara, 2008), then
they would be expected to have relatively depleted trace element compositions, but this is
not the case (Janney and Castillo, 2001). Although MORB from some hotspot-influenced
ridge segments have systematically lower Na8 , Ti8 and higher Fe8 (section 3.6), our samples from cores in the oldest Atlantic crust do not have elevated La/Sm ratios, as would
be expected for lavas erupted at ridges close to hotspots (section 3.6). These drillsites are
located over a wide area, and well away from former positions of known ‘hotspots’. These
observations suggest that changes in mantle source composition are not responsible for
the temporal changes in the major element chemistry of Atlantic MORB. An important
constraint on the origin of the systematic temporal variations in MORB composition is
Dissertation P.A. Brandl
23
3. High mantle temperatures following rifting
that these are observed within the Atlantic, and probably the Indian Ocean (Humler
et al., 1999; Humler and Besse, 2002), but not within the Pacific. During the Mesozoic,
spreading ridges in the Atlantic and Indian Oceans were located close to rifted continental margins, whereas the oldest Pacific lavas at Site 801 were erupted >2,000 km from
the nearest continental margin (Fisk and Kelley, 2002). We therefore suggest that the
temporal variations within Atlantic MORB result from long-term changes in the thermal
structure of the upper mantle related to continental rifting.
Figure 3.3: Major element compositions of MORB from spreading ridges in young ocean basins.
Variation of (a) Na8 and (b) Fe8
with spreading ridge water depth for
MORB from spreading centres in the
Red Sea/Gulf of Aden (red symbols),
northern Central Indian Ridge (yellow), southern Central Indian Ridge
(blue), and zero-age Indian Ocean
MORB from other spreading centres
(grey). Spreading centres formed in
young ocean basins shortly after continental rifting (Red Sea; 5 Ma) lie at
shallow water depths and erupt MORB
with low Na8 , high Fe8 , whereas lavas
from ridges in ocean basins associated
with mature passive margins (southern
Central Indian Ridge; 115 Ma) have
higher Na8 and lower Fe8 .
This interpretation is supported by the compositions of zero-age MORB erupted in regions of recent continental breakup. Lavas from active spreading centres in the Red Sea,
where spreading began at 5 Ma following rifting between the African and Arabian Plates,
have relatively low Na8 and high Fe8 , similar to the oldest drilled samples from the Atlantic (Fig. 3.3, and Ligi et al., 2012). MORB from the southern Central Indian Ridge
and other mature Indian Ocean spreading centres have higher Na8 and lower Fe8 , similar
to those of zero-age Atlantic MORB. Active spreading ridges in the Gulf of Aden and
northern Central Indian Ridge, where the oldest seafloor is 15–50 Ma in age, erupt lavas
with intermediate compositions (Fig. 3.3). The fractionation-corrected compositions of
ancient Atlantic and Indian MORB and zero-age lavas from active spreading centres in
the Indian Ocean, Gulf of Aden and Red Sea, vary with the age difference between the
eruption ages of the lavas and the time of local continental breakup (Fig. 3.4). For an age
difference of <60-70 Ma, almost all MORB irrespective of their eruption age, have higher
24
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
Fe8 and lower Na8 than the median of zero-age MORB.
Figure 3.4: Evolution
of
mantle temperature with
time following continental
breakup. Variation of mantle
potential temperature (Tpot )
estimated using (a) Na and (b)
Fe (Kelley et al., 2006), with the
difference between eruption age
and the time of local continental
breakup (δAge). Black triangles;
ancient Indian Ocean MORB
(PetDB), other symbols as in
Figure 3.1. Temperatures inferred
from ancient and zero-age MORB
erupted at spreading ridges
soon after their initial development following continental
rifting and breakup are up to
about 150◦ C higher than the
median for zero-age Atlantic and
Indian Ocean MORB (yellow
field represents one standard
deviation) from spreading ridges
which have been in existence
since continental breakup in the
Mesozoic (histogram).
Humler and Besse (2002) observed a correlation between the major element composition of zero-age MORB from the Atlantic and Indian Oceans and the distance to the
nearest passive margin, and interpreted this as an effect of ‘thermal transfer’ between
continents and the upper mantle beneath ocean basins. We suggest that the apparent
cooling of the mantle beneath the Central Atlantic over the past 170 Ma is the result
of removing the insulating effect of continental lithosphere, following continental breakup
200 Ma ago. Continental lithosphere, through which heat must be transported by conduction, impedes heat loss from the mantle. This ‘continental insulation’ effect is predicted
to lead to warming of the underlying mantle (Grigné and Labrosse, 2001; Lenardic et al.,
2005; Zhong and Gurnis, 1993; Coltice et al., 2007). The precise magnitude of the warming is debated, because the insulation effect is partly offset by increased convective vigour
due to the lower mantle viscosity at higher temperature (Lenardic et al., 2005), but it is
estimated that supercontinent formation may result in mantle warming of around 100◦ C
in 100 Ma, due to the combined effects of insulation and longer-wavelength mantle flow
(Coltice et al., 2007, 2009).
Dissertation P.A. Brandl
25
3. High mantle temperatures following rifting
Our data can be used to quantify the approximate magnitude of mantle warming resulting from continental insulation. We calculated primary magma temperatures using
three different thermometers (see section 3.6). Assuming that differences in fractionationcorrected Na2 O, FeO and MgO result from mantle temperature variations (Klein and
Langmuir, 1987), yields temperatures (Kelley et al., 2006; Herzberg et al., 2007) for the
oldest drilled lavas from the Atlantic and Indian Oceans that are up to about 150◦ C higher
than the average for zero-age MORB (Fig. 3.4, see section 3.6). In the Atlantic, the thermal anomaly apparently persisted for approximately 60–70 Ma after continental rifting
(Fig. 3.4). This timescale is consistent with that predicted by models (Rolf et al., 2012;
Coltice et al., 2009), and similar to the age range of anomalously smooth, thick crust in
the Atlantic and Indian Oceans previously attributed to the continental insulation effect
(Whittaker et al., 2008).
Our data show that the average temperature of the upper mantle evolves for several
tens of Ma after continental rifting and development of new spreading centres. This effect has implications for past seawater composition, volcanic CO2 output and sealevel,
which must be taken into account in models which attempt to reconstruct these parameters through time (Gaffin, 1987; Berner and Kothavala, 2001; Hardie, 1996). For
example, the average ridge depth inferred from the Na8 of 160 Ma Atlantic MORB is
750–1,000 m, whereas the average depth of the Mid-Atlantic ridge system today is
2,900 m. Variations in the average depth of former spreading centres, as well as changes in
the area-age distribution of seafloor (Müller et al., 2008b), need to be taken into account
in order to calculate accurately changes in sealevel resulting from plate tectonic processes
over geological timescales.
3.5 Acknowledgements
This research used samples provided by the Integrated Ocean Drilling Program (IODP).
Funding for this research was provided by the Deutsche Forschungsgemeinschaft (grants
RE3020/1-1 and 1-2), and P.A.B. acknowledges a doctoral fellowship of the Erika Gierhl
Foundation. We thank A. Richter, H. Brätz, C. Kaatz, N. Hohmann, F. Stöckhert, C. Weinzierl, F. Genske, and the curators at Bremen and Gulf Coast Core Repositories for their
help during sampling and data analysis, and the Smithsonian Institution for providing
MORB glass standards.
26
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
3.6 Supplementary Information
3.6.1 Sampling and analytical methods
We personally sampled 45 DSDP-ODP-IODP sites (12 in the Atlantic and 33 in the Pacific) drilled into normal oceanic crust formed at spreading centres (no intraplate lavas)
ranging in age from 6 up to 170 Ma. We selected only glassy pillow rinds showing no or
only minor visible alteration (244 fresh glasses from the Pacific and 298 from the Atlantic,
since the Atlantic sites contain much more glasses relative to core length). Age information for individual sites were taken from the Initial Reports of the DSDP, ODP and
IODP Proceedings or (when no precise age information was given in the Initial Reports)
from the digitised age and spreading rate grids of Müller et al. (2008a) and Müller et al.
(2008b) available through geomapapp (Ryan et al., 2009; http://www.geomapapp.com).
An overview of drillsite locations and the age of the oceanic crust can be found in Figure
3.5. A short summary of sampled sites (geographic position, waterdepth, depth to basement, number of fresh glass samples analysed) is given in Table A1 of the Appendix.
Figure 3.5: Global map of the age of the oceanic crust processed using GMT (Wessel and Smith,
1991, 1995) showing the locations of DSDP-ODP-IODP sites sampled in our study. Colours correspond to
the age of the oceanic crust as inferred from magnetic lineations (Müller et al., 2008a,b). Black diamonds
represent sites located on ‘normal’ oceanic crust, and red diamonds represent sites that are influenced
by mantle material distinct from normal MORB mantle (Azores, MAR at 45◦ N). Location of previously
analysed samples of ancient MORB are shown by black triangles. White circles show locations of zero-age
MORB samples. Abbreviations: JdF – Juan de Fuca Ridge, EPR – East Pacific Rise, GSC – Galapagos
Spreading Centre, PAR – Pacific-Antarctic Ridge, MAR – Mid-Atlantic Ridge, CIR – Central Indian
Ridge, SWIR – Southwest Indian Ridge, SEIR – Southeast Indian Ridge.
Dissertation P.A. Brandl
27
3. High mantle temperatures following rifting
Major element compositions of glasses were determined using a JEOL JXA-8200 Superprobe electron microprobe at the GeoZentrum Nordbayern (Friedrich-Alexander-Universität Erlangen-Nürnberg), with 15 kV acceleration voltage, 15 nA beam current and a
beam diameter of 10 µm (for more details see Brandl et al., 2012). Data in Tables A2
and A3 of the Appendix (Atlantic and Pacific, respectively) represent averages of 10 individual analyses on single glass chips. Of the samples analysed, 340 from 30 sites showed
no evidence of seawater alteration, as inferred from major element totals (97.5 wt. %
or higher), and K2 O concentrations (depending on MgO content but generally below
0.2 wt. % in MORB; Tables A2 and A3, Appendix).
Natural glass standards VG-A99 and VG-2 of the Smithsonian Institution (Jarosewich
et al., 1980 with S and Cl data of Thordarson et al., 1996 and Jenner and O’Neill, 2012)
were analysed with each set of samples in order to monitor analytical precision and accuracy (Table A4, Appendix). The raw data listed in Tables A2 and A3 (Appendix) have
been normalised to values for the VG-2 standard (see Table A4, Appendix: VG-2 preferred value) of SiO2 50.48, TiO2 1.86, Al2 O3 14.03, FeOT 11.76, MnO 0.21, MgO 6.82,
CaO 11.08, Na2 O 2.64, K2 O 0.19, P2 O5 0.21 (all in wt. %), S 1394 ppm and Cl 302 ppm,
which yielded the following values for VG-A99: SiO2 51.20, TiO2 4.11, Al2 O3 12.49, FeOT
13.35, MnO 0.18, MgO 4.96, CaO 9.30, Na2 O 2.71, K2 O 0.85, P2 O5 0.44 (all in wt. %),
S 145 ppm and Cl 240 ppm.
After removal of the carbon coating used for microprobe analysis, the Rare Earth Element (REE) concentrations of selected samples were determined using laser-ablation
ICPMS (Indouctively Coupled Plasma Mass Spectrometry), on the same glass chips used
for microprobe analysis. Measurements were conducted on the LA-ICPMS system at
the GeoZentrum Nordbayern in Erlangen (New Wave Research UP193FX laser ablation
system coupled to an Agilent 7500i quadrupole ICPMS). Measurement conditions were
0.75 GW cm−2 laser energy and 3.74 J cm−2 energy density on 50 µm spots. Precision
and accuracy (generally better than 10 %) were checked by repeated measurements of the
secondary standards NIST-614 and BCR-2G.
3.6.2 Zero-age MORB reference database
We compiled a reference dataset of major element data for zero-age MORB glasses using
the PetDB database (http://www.petdb.org), together with data from the Smithsonian
Abyssal Volcanic Glass Data File (AVGDF; http://mineralsciences.si.edu/research/glass/
vg web.xls). Locations of the zero-age MORB glasses in this dataset are shown in
Figure 3.5.
28
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
Figure 3.6: SiO2 , TiO2 , Al2 O3 , FeOT , CaO, Na2 O and K2 O versus MgO (all concentrations
in wt. %). Zero-age MORB glasses shown by grey dots, and ancient MORB by coloured triangles
(literature) and diamonds (this study). Colours indicate the age of the drilled samples (see colour bar).
Ancient Pacific MORB generally display a larger variability in MgO (degree of fractionation) than ancient
Atlantic MORB, a feature which is also observed in zero-age MORB from the Pacific and Atlantic. Ancient
MORB samples lie within the compositional range of modern MORB, and the average MgO values of
both are similar. The fractionation correction should therefore not lead to a systematic chemical bias
between the ancient and zero-age datasets, since the magnitude of the correction for both is similar.
Dissertation P.A. Brandl
29
3. High mantle temperatures following rifting
The initial selection criteria for samples from the PetDB database were: (a) Class: Igneous, volcanic, mafic including alkali basalt, basalt, basaltic breccia, basaltic rubble,
picrite, tholeiite and trachybasalt, (b) Alteration: Fresh only, (c) Spreading centres: All,
(d) Glass data only, (e) Precompiled. To this compilation we added the Smithsonian
AVGDF (with tectonic code ‘R: spreading ridges’ and with age ≤1 Ma only), filtered for
duplicates, excluded data lacking water depth information or major element data, those
from spreading ridges shallower than 400 m (likely influenced by hotspots), and those
samples with SiO2 concentrations <45 or >53 wt. % (similar to Niu and O’Hara, 2008)
or K2 O concentrations higher than the low-K tholeiitic series defined in Rickwood (1989).
We separated these data (>9,800 individual samples in total) into three groups according
to their location in the Atlantic, Indian or Pacific Ocean. Our ancient Indian MORB
database (Fig. 3.4) was treated in the same way as other published and compiled data
to exclude highly alkaline lavas erupted during the early stages of rifting in the northern
Red Sea. Other ancient MORB from the Atlantic and Pacific Oceans also do not include
samples with K2 O >0.2 wt. % except for some basaltic andesites from IODP Hole 1256D,
which result from high degrees of fractional crystallisation (SiO2 >52 wt. %).
Figure 3.7: Fractionationcorrected Na2 O (Na8 ) and
FeOT (Fe8 ) values of zero-age
(grey points) and ancient MORB
(symbols as in Fig. 3.6; literature
data for ancient Indian MORB
represented by black triangles).
Averages of zero-age MORB from
500 m ridge depth intervals (from
500 to 5,500 m waterdepth; Table
A5, Appendix) are shown for
comparison (hexagons).
The major element analyses contained in our zero-age MORB database were obtained in
many different laboratories, using different analytical methods and are reported relative
to different standards. Some studies do not report any standard data at all. As far as
this was possible, we have corrected the zero-age MORB dataset for the effects of interlaboratory bias. In particular, we have normalised our data (see above) and data from
the Smithsonian laboratories (approximately 5,900 samples), which apparently have lower
MgO and higher SiO2 , Al2 O3 and CaO than data obtained elsewhere (Langmuir et al.,
1992), to the preferred values of the VG-2 standard listed in Table A4 (Appendix). The
corrected major element compositions of zero-age MORB and ancient (drilled) MORB
30
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
samples from the Atlantic and Pacific Oceans are shown in Figure 3.6.
3.6.3 Correction for the effects of fractional crystallisation
In order to identify chemical variations resulting from source composition and melting
processes, it is essential to correct samples in both datasets for the effects of fractional
crystallisation. Since the samples from many of our sites do not display a wide enough
range in fractionation to estimate liquid lines of descent for individual sites (Fig. 3.6),
we used previously published correction methods (Klein and Langmuir, 1987; Taylor and
Martinez, 2003; Kelley et al., 2006) to calculate the concentrations of Na2 O, FeOT and
TiO2 of melts at a MgO concentration of 8.0 wt. %, using a simple linear regression.
We corrected only samples having a MgO concentration greater than 7.0 wt. % (>7,500
samples of our zero-age MORB dataset) in order to minimise errors produced by the
correction method itself (Kelley et al., 2006; Niu and O’Hara, 2008). The same MgO filter
was used for both the modern and ancient MORB datasets, to avoid possible bias due to
differences in the magnitude of the correction for each dataset. The averages from 500 m
ridge depth intervals (from 500 to 5,500 m waterdepth) are shown in Figure 3.7 and in
Table A5 (Appendix), together with fractionation-corrected data (Table A6, Appendix).
3.6.4 Calculation of primary magma compositions
An estimate of the primary magma composition is necessary for estimating mantle potential temperatures. Our best estimate of primitive magma composition is based on the
partitioning between Mg# of the liquid (glass) and the forsterite content of olivine, following the work by Roeder and Emslie (1970). The Fo in primitive MORB olivines is
between 89 (Niu and O’Hara, 2008), 90 (Kelley et al., 2006) and 91.5 (Siqueiros Transform; Perfit et al., 1996), which means that the Mg# of the primitive liquid must be
in the range of 68 to 78 depending on the assumed Fo value of primitive olivine and
Kol–liq
D(Fe-Mg) = 0.30±0.03 (Roeder and Emslie, 1970; Putirka, 2008b).
Mineral-liquid (or mineral-glass) thermobarometry (such as olivine-liquid or clinopyroxeneliquid) provides an accurate estimate of the crystallisation temperature of melt and crystal only when the phases in question are in equilibrium (e.g., Putirka, 2008b). Glass or
whole-rock compositions must either lie on the olivine control line or must be traced to it
(Putirka, 2005), which is an additional requirement aside of mineral and liquid being in
equilibrium. Most of our samples have fractionated not only olivine, but also other phases
(see Fig. 3.8). Thus, it is necessary to correct our glass compositions for the effects of fractional crystallisation as far as the olivine control line (≥8.5 wt. % in Kelley et al., 2006 or
≥9.5 wt. % in Putirka, 2008a), where FeOT is almost constant at variable MgO (Fig. 3.8a).
Dissertation P.A. Brandl
31
3. High mantle temperatures following rifting
In order to calculate primary magma compositions, we used the method of Kelley et al.
(2006), who improved and extended the original fractionation correction method of Klein
and Langmuir (1987) for TiO2 and H2 O. Kelley et al. (2006) calculated the correction in
three distinct steps in order to account for the observed changes in slope of the liquid line
of descent (LLD) between 8.0 and 8.5 wt. % MgO in MORB suites (Fig. 3.8a), which result from changes in the crystallising assemblage. The equations for the correction method
for data with MgO <8.5 wt. % (two steps between 7.0 and 8.0 and 8.0 to 8.5 wt. % MgO,
respectively) can be found in Kelley et al. (2006).
After these corrections were applied, or when the MgO concentration in the sample
exceeded 8.5 wt. %, a third incremental step was applied to calculate primary magma
compositions that could be used for thermometric calculations. Important assumptions
for this third correction step include: (a) initial Fe2+ /ΣFe of 0.9 (Niu and O’Hara, 2008),
(b) olivine as the only phase fractionating from the liquid (‘olivine control line’), (c) equilibrium fractionation of olivine, (d) a mantle olivine composition of Fo = 90 (Kelley et al.,
2006), and (e) Kol–liq
D(Fe-Mg) = 0.30 (Kelley et al., 2006; Putirka, 2008b).
Figure 3.8: MgO versus (a)
FeOT and (b) CaO for ancient MORB glasses. The
mean MORB olivine control line
(black dashed line) of Putirka
et al. (2007) is shown for comparison. Most of our samples
show evidence of having crystallised other phases (clinopyroxene, plagioclase, spinel) in addition to olivine (i.e. samples do
not lie on the olivine control line;
change in MgO-CaO slope indicated by the black arrow in (b)).
More fractionated samples with
MgO <7.0 wt. % have been excluded from fractionation correction (grey field). The slope of the
fractionation correction method
of Kelley et al. (2006) is shown
with a yellow line in (a), together
with the minimum MgO concentration at which olivine is the only
phase crystallising from the liquid as estimated by Kelley et al.
(2006) (red) and Putirka et al.
(2007) (blue).
32
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
We calculated the Mg# of the liquid and the Fo content of olivine in equilibrium with
the respective liquid. We then added olivine to the liquid in 0.5 % steps (at each step
recalculating the composition of olivine in equilibrium with the liquid) and recalculated
the liquid composition in regard of a change in Fe2+ /ΣFe, assuming that olivine almost
completely excludes Fe3+ (O’Neill et al., 1993). The total amount of olivine that was required to be added to the liquid was about 5–17 % for our ancient MORB samples. The
resulting primary magma MgO concentrations are typically 11–14 wt. % and thus in the
range for primary (N-)MORB melts (10–13 wt. %) reported by Herzberg et al. (2007).
Our results for primary FeOT concentrations are in good agreement with the range of 7.0
to 9.6 wt. % (of which 95 % of MORB are fractionated) reported by Putirka (2008b).
We applied the same correction methods to both the zero-age and ancient MORB datasets.
About 7,500 samples of our zero-age MORB database have MgO higher than 7.0 wt. %
and we have therefore calculated both primitive (corrected for the effects of fractional
crystallisation to MgO = 8.0 wt. %) and primary magma compositions (melt composition
in equilibrium with Fo = 90) for these samples.
3.6.5 Comparison of ancient and zero-age MORB datasets
Sampling sites for zero-age MORB along the active mid-ocean ridge system are not equally
distributed and some regions are much more densely sampled than others. As a result,
the zero-age MORB reference database we have compiled using PetDB is unlikely to yield
an accurate estimate of the mean composition and variability of average global zero-age
MORB. Given their relative lengths, spreading ridges in the Atlantic and Indian Oceans
may be under-represented in our zero-age MORB database, compared to Pacific ridges.
Since we consider Pacific and Atlantic MORB separately, this should not matter for our
purposes provided that the Pacific and Atlantic samples in our database are representative
of ridges in the respective oceans. This is unlikely to be the case however; for instance in
the Pacific, the East Pacific Rise is far more densely sampled than the Pacific-Antarctic
Rise. Niu and O’Hara (2008) compiled a similar database of global MORB compositions
(N = 9,130) and we have compared our dataset to theirs to check for differences (Fig. 3.9)
that could arise from sampling bias.
The distributions of the sampled ridge depth curves of our zero-age MORB dataset (including Indian ridges) and that of Niu and O’Hara (2008) are similar (Fig. 3.9), and
the mean sampled waterdepth in the Niu and O’Hara (2008) database (2,803 m), is very
similar to ours (mean 2,832 m, median 2,677 m). This depth is slightly greater than the average depth of the mid-ocean ridge system (2,600 m) estimated by Stein and Stein (1992).
Nevertheless, it is likely that our zero-age Pacific MORB dataset is biased towards shallow
Dissertation P.A. Brandl
33
3. High mantle temperatures following rifting
Figure 3.9: Frequency of sampled spreading ridge water depths for the Niu and O’Hara (2008)
zero-age MORB compilation (grey), and our global zero-age MORB compilation, including all samples
(green), those from Pacific spreading ridges (blue), and Atlantic and Indian spreading ridges (orange).
depths (and thus lower Na8 ), because of the relatively large number of samples from the
Galapagos Rise Spreading Centre. Since the major element chemistry of MORB varies
systematically with spreading ridge depth (Fig. 3.10), it should be possible to calculate
an ‘average global MORB composition’ from the depth profile of the global MOR system.
Figure 3.10: Variation of Na8 and Fe8 with spreading ridge depth for global zero-age MORB.
Samples were sorted into 500 m depth intervals according to their eruption depth, and Na8 and Fe8 values
calculated for all samples within each depth interval then averaged (see Table A5, Appendix).
Using the relationships in Figure 3.10, a mean MOR depth of 2,600 m would correspond
to a Na8 value of 2.48 wt. %, whereas a ridge depth of 2,832 m corresponds to a Na8
of 2.55 wt. %. This difference (0.07) is small compared to the total range in Na8 in Pacific and Atlantic MORB (approximately 2.0), and an order of magnitude smaller than
the difference between ancient and zero-age Atlantic MORB (about 0.9). We therefore
34
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
believe that our zero-age MORB dataset is sufficiently representative of global MORB
compositions.
As we are dealing with sample sets that are limited in sample size and distribution,
statistical tests of the significance of differences between the zero-age and ancient MORB
datasets are very limited. We tested our data for normal distribution but for most chemical
parameters the results were negative indicating no normal distribution. Thus, statistical
tests for differences between these datasets such as the t-test are meaningless since a normal distribution is a prerequisite for these tests. The group size of our ancient MORB
database is small (30 distinct sites) and drillsites are not uniformly distributed on ancient
oceanic crust. It is therefore possible that the chemical differences we observe between
Figure 3.11: Comparison
of data of modern MORB
(PetDB) and ancient MORB
from the Central Atlantic.
Blue graph displays the present
day water depth variation along
the median rift valley of the MidAtlantic Ridge (MAR) between
10◦ and 30◦ N (inferred from
Google Earth) with the general
trend shown by the blue shaded
field. Grey symbols show Na8
values of zero-age MORB from
the MAR axis. DSDP-ODP sites
have been projected back along
flowlines back to the position
where they would be located at
the present-day MAR. Greatest
offset for oldest sites (blue),
smaller for the youngest site 396
(red).
ancient and zero-age MORB could be an artefact of regional chemical variations. Since
many of the oldest Atlantic drill sites are located in the Central Atlantic, we compared
the compositions of ancient MORB from this region with the chemical variation observed
in zero-age MORB from the Central Mid-Atlantic Ridge only (Fig. 3.11), in order to test
for possible geographic bias effects.
From Figure 3.11 it is apparent that ancient Atlantic MORB have significantly lower
Na8 than zero-age MORB erupted along the MAR at the same latitude of eruption, as
estimated by projecting the drill site locations along flowlines to the present-day MAR.
Na8 values as low as those of lavas from Atlantic Sites 105 and 534A are extremely rare
Dissertation P.A. Brandl
35
3. High mantle temperatures following rifting
in zero-age MORB from the Central North Atlantic between 10 and 30◦ N (Fig. 3.11).
Most of the drillsites we sampled penetrated less than 200 m into igneous basement
(Fig. 3.12). Lavas collected by drilling into the uppermost parts of the mature oceanic
crust will represent the youngest flows erupted at that location. In contrast, zero-age
MORB samples are generally dredged from the axial region of active spreading ridges,
and these will eventually make up the lowermost parts of mature oceanic crust. If significant chemical differences exist between flows erupted on- and off-axis, then the observed
chemical differences between drilled and dredged samples could arise from differential
sampling of the upper (younger) and lower (older) parts of the extrusive section of the
oceanic lithosphere.
Figure 3.12: Age of oceanic
crust versus penetration into
the igneous oceanic crust.
Most sites that have penetrated
the igneous oceanic crust have
been terminated after <200 m of
drilling into basement.
To test for possible differences in composition between flows erupted on- and off-axis, we
compiled major element data for samples from the 9–12◦ N region of the East Pacific Rise,
for which the precise eruption location is known from mapping and radiometric dating
(Sims et al., 2003; Turner et al., 2011; Waters et al., 2011). These data show that offaxis flows tend to have lower MgO, and lower Na2 O for a given MgO, compared to lavas
erupted at the ridge axis (Fig. 3.13). Thus, if the EPR is representative of other parts
of the mid-ocean ridge system, the uppermost, off-axis flows preferentially sampled by
drilling would tend to have higher Na8 , whereas low Na8 are observed in ancient Atlantic
crust. In addition, if a difference in composition existed generally between the upper and
lower parts of the oceanic crust, we would expect to see a systematic difference between
zero-age MORB and all drilled, ancient MORB samples, which is not the case.
36
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
Figure 3.13: MgO
versus
T
TiO2 , FeO
and Na2 O of
zero-age Pacific MORB (grey
symbols), together with lava
flows from the East Pacific Rise
at 9–12◦ N which are known to
have been erupted either on-axis
(yellow stars) or off-axis (red
stars). EPR data from Sims et al.
(2003), Turner et al. (2011) and
Waters et al. (2011).
3.6.6 Calculation of mantle potential temperatures
Thermometry of MORB has been a controversial topic over the past decades (e.g., Klein
and Langmuir, 1987; Langmuir et al., 1992; Herzberg and O’Hara, 2002; Herzberg et al.,
2007; Niu and O’Hara, 2008). Nevertheless, recent studies (e.g., see review paper by
Putirka, 2008b and references therein) provide powerful tools to estimate magmatic temperatures from mineral-liquid equilibria or liquid compositions only. We calculated mantle
temperatures using the Na and Fe (Kelley et al., 2006) and Mg (Herzberg and O’Hara,
2002) thermometers, together with the estimated primary magma composition for all
samples. Although there is some uncertainty in the absolute temperature calculated using
these different methods, the relative temperature differences between samples calculated
with the same method are considered to be robust.
The use of Fe, Mg and Na to calculate mantle temperatures requires the assumption
that significant major element heterogeneity does not exist in the upper mantle source of
MORB, which is unlikely to be the case (e.g., Niu and O’Hara, 2008). Na is probably more
sensitive than Fe or Mg to variations in mantle composition. However, over the expected
temperature range of 1,250–1,400◦ C, the variation of Na concentration in primary melts is
Dissertation P.A. Brandl
37
3. High mantle temperatures following rifting
Figure 3.14: Mantle potential temperature (Tp , in ◦ C) calculated from Fe and Na (Kelley
et al., 2006) in inferred primary magmas of MORB from the Atlantic and Indian Oceans (red) and Pacific
(blue). Note that the difference in mean temperature obtained using the different thermometers (15–20◦ C)
is similar to the temperature difference between ocean basins (25–30◦ C), and small compared to both
the inferred range in temperature beneath each ocean basin (approximately 200◦ C), and the apparent
temporal change in temperature of the Atlantic upper mantle as inferred from Na (approximately 150◦ C).
Temperatures given in this figure correspond to median mantle potential temperatures (with one standard
deviation) for Atlantic and Indian Ocean (red) and Pacific (blue).
also about 4 times greater than either Fe or Mg, and so in the main text we have focused
on temperatures calculated using the Na thermometer.
Mean mantle potential temperatures calculated for zero-age Atlantic MORB unfiltered
for possible hotspot effects are 1,375±45◦ C and 1,395±55◦ C for the Fe and Na thermometers respectively, whereas for our Pacific zero-age MORB dataset we obtained temperatures of 1,405±40◦ C and 1,420±35◦ C (Fig. 3.14). Excluding ridges that are influenced by
hotspots (such as Reykjanes and Kolbeinsey Ridges, the Mid-Atlantic Ridge south of the
Azores, and the Galapagos Spreading Centre) yields slightly lower (global) mantle potential temperatures of 1,385±40◦ C (Mg), 1,390±45◦ C (Fe) and 1,400±40◦ C (Na). These
temperatures are in good agreement with other published estimates of the temperature of
the upper mantle beneath spreading ridges: e.g., 1,300–1,570◦ C (Langmuir et al., 1992),
1,370±50◦ C (Putirka, 2005) and 1,280–1,400◦ C (Herzberg et al., 2007).
The Fe and Mg thermometers yield similar temperatures for MORB older than 150 Ma in
the Pacific and Atlantic (1,420–1,465◦ C), which are higher than those of zero-age MORB.
Temperatures derived using the Na thermometer for ancient Atlantic MORB are significantly higher (approximately 1,510◦ C) than Pacific MORB of the same age, and zero-age
MORB from either ocean basin. The calculated differences in mantle potential temperature between ancient and zero-age MORB are thus greater in the Atlantic than in the
Pacific. Similarly low Na8 and high inferred mantle temperatures characterise zero-age
MORB from the Red Sea and Gulf of Aden, which like the ancient drilled Atlantic and
38
Dissertation P.A. Brandl
3. High mantle temperatures following rifting
Indian MORB were erupted within 25 Ma of continental rifting and breakup (see Fig. 3.3
and 3.4). Our preferred explanation for higher mantle temperatures in ancient Atlantic
MORB is therefore that they result from continental insulation effects, as discussed in the
main text.
3.6.7 Effects of mantle heterogeneity and hotspots
Klein and Langmuir (1987) showed that MORB from active spreading ridges in the vicinity of hotspots tend to have lower Na8 for a given Fe8 . The lower Na8 of ancient Atlantic
MORB could therefore result from the thermal effect of nearby hotspots, or from melting
of chemically distinct hotspot mantle.
Figure 3.15: Fractionation corrected Na8 and Fe8 values for zero-age MORB lavas from spreading ridges that are clearly not influenced by hotspots (light grey symbols) and MORB from ridges in the
vicinity of hotspots (black), including the Iceland region (Reykjanes and Kolbeinsey ridges), the MidAtlantic Ridge south of the Azores and the Galapagos Spreading Centre. Coloured hexagons represent
depth-averaged zero-age MORB compositions (Table A5, Appendix). All data are from our zero-age
MORB compilation. The mean compositions of ridge segments that were interpreted by Klein and Langmuir (1987) to be influenced by hotspots are shown by yellow circles (enclosed by yellow shaded field).
Data for ancient MORB (coloured triangles and diamonds) are shown for comparison.
Most ancient Atlantic MORB lie at the low Na8 and high Fe8 end of the global zeroage MORB array (Fig. 3.15). In contrast, MORB from regions influenced by hotspots
display relatively low Na8 at a given Fe8 . Most Atlantic MORB from ridge segments
close to hotspots also have distinct trace element compositions to normal MORB far
from hotspots. For example, zero-age Atlantic MORB from the vicinity of Iceland, the
Azores, Ascension, St. Helena and Tristan da Cunha have chondrite-normalised La/Sm
ratios (La/Sm)N that exceed 1, reaching up to 3 (e.g., Schilling et al., 1983; Fontignie and
Dissertation P.A. Brandl
39
3. High mantle temperatures following rifting
Schilling, 1996). We measured La/Sm ratios of representative ancient Pacific and Atlantic
MORB from the drillsites we studied, using laser-ablation ICPMS to analyse the same
glass chips used for major element analysis (Fig. 3.16).
Figure 3.16: Chondritenormalised (La/Sm)N (McDonough and Sun, 1995) and
age (Ma) of selected ancient
drilled MORB samples measured
using laser-ablation ICPMS.
Almost all glasses analysed
are N-MORB according to the
classification of Schilling et al.
(1983). Ancient Pacific MORB
(blue) include some transitional
MORB with (La/Sm)N between
0.7 and 1.8, but ancient Atlantic
glasses (red) are exclusively
N-MORB. Data from Brandl et
al. (manuscript in preparation).
Although some Pacific samples have (La/Sm)N ratios of between 0.7 and 1.0, all ancient
Atlantic MORB have (La/Sm)N <0.7 (Fig. 3.16), and lie within the field of ‘normal’
depleted (N-)MORB as defined by Schilling et al. (1983). There is therefore no evidence
from the trace element compositions of ancient Atlantic MORB for an influence from
chemically anomalous hotspot mantle.
Janney and Castillo (2001) showed that ancient Atlantic MORB have similar or slightly
enriched isotope compositions compared to most zero-age MORB away from hotspots,
and that the isotope variations within the older samples cannot be explained by mixing
of depleted Atlantic MORB mantle with any known Central Atlantic hotspot (e.g., Cape
Verde). In addition, all drillsites in ancient Atlantic oceanic crust are located away from
seamount chains, seafloor swells, and the inferred location of presently-active Atlantic
‘hotspots’. We therefore believe that the distinct major element compositions of ancient
Atlantic MORB do not result from hotspot influence.
40
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
4 Volcanism on the flanks of the East
Pacific Rise: quantitative constraints
on mantle heterogeneity and melting
processes
Philipp A. Brandl1 , Christoph Beier1 , Marcel Regelous1 , Wafa Abouchami2,3 , Karsten
M. Haase1 , Dieter Garbe-Schönberg4 and Stephen J.G. Galer3
1
GeoZentrum Nordbayern, Friedrich-Alexander-Universität Erlangen-Nürnberg, Schlossgarten 5, 91054 Erlangen, Germany
2
Institut für Mineralogie, Westfälische Wilhelms-Universität Münster, Corrensstr. 24,
48149 Münster, Germany
3
Max Planck Institute for Chemistry, P.O. Box 3060, 55020 Mainz, Germany
4
Institut für Geowissenschaften, Christian-Albrechts-Universität Kiel, Ludewig-Meyn-Str.
10, 24118 Kiel, Germany
Abstract
We present major and trace element and Sr, Nd and triple-spike Pb isotope data for
17 fresh volcanic glasses from Seamount 6, a 10-km diameter seamount located 140 km
east of the East Pacific Rise (EPR) at 12◦ 45’N. Geological and geochronological evidence show that magma compositions evolved from tholeiitic basalts to alkalic basalts and
basaltic trachyandesites during the 1–2 Ma active lifetime of the seamount. Major and
trace element compositions in Seamount 6 lavas vary systematically with isotope ratios;
the youngest lavas with the highest incompatible trace element concentrations have the
highest La/Yb, Nb/Zr, K2 O/TiO2 , 87 Sr/86 Sr, 206 Pb/204 Pb and the lowest 143 Nd/144 Nd,
MgO and CaO. The range in element concentrations, incompatible element ratios, and
isotope compositions in Seamount 6 lavas exceeds that observed in lavas erupted at the
adjacent ridge axis, and is comparable to the range in lava compositions reported from all
near-ridge seamounts studied to date. The observed range in lava compositions is consistent with mixing between enriched and depleted melts at shallow levels in the crust. The
Dissertation P.A. Brandl
41
4. Volcanism on the flanks of the East Pacific Rise
inferred difference in composition between these mixing endmembers cannot be explained
by variable degrees of melting of a single source composition, and requires that the upper
mantle is extremely heterogeneous on the scale of the melting region beneath a single
seamount.
We can show that the range in composition of EPR seamount lavas cannot be generated
by melting of variably heterogeneous mantle in which enriched and depleted materials
contribute equally to melting (source mixing). Instead, the trace element and isotope
compositions of seamount lavas can be reproduced by melting models in which more enriched, fertile mantle lithologies are preferentially melted during mantle upwelling. At
progressively lower degrees of melting, erupted lavas are thus more enriched in incompatible trace elements, have higher La/Yb, K/Ti, 87 Sr/86 Sr ratios and lower 143 Nd/144 Nd. If
this is a common process, then mantle-derived magmas are unlikely to inherit the average incompatible trace element and isotope composition of their mantle source, which is
likely to be significantly more depleted, nor will they display the full range of compositions
present in the mantle melting region. These results have implications for the way in which
oceanic basalts can be used to infer the composition of the upper mantle.
4.1 Introduction
At mid-ocean ridges, plate separation and mantle upwelling result in mantle melting and
eruption of approximately 3 km3 mid-ocean ridge basalt (MORB) per year (Crisp, 1984).
The chemical and isotopic compositions of MORB are commonly used to infer the composition of the upper mantle, and the processes of mantle melting, melt transport and
crystallisation. Numerous geochemical studies of MORB (e.g., Allègre and Turcotte, 1986;
Arevalo Jr. and McDonough, 2010; Donnelly et al., 2004; Hirschmann and Stolper, 1996;
Niu et al., 1999; Phipps Morgan and Morgan, 1999; Schilling et al., 1983; Zindler and
Hart, 1986) have shown that even in the absence of nearby hotspots, the upper mantle
is chemically and isotopically heterogeneous. The ultimate origin of this heterogeneity
is unclear, but likely results from subduction and ‘recycling’ of oceanic crust, sediment,
metasomatised oceanic lithosphere, mantle wedge material, or oceanic island chains and
plateaus (Donnelly et al., 2004; Niu and O’Hara, 2003; Pilet et al., 2005; White and Hofmann, 1982).
Most MORB are apparently derived from mantle that is depleted in highly incompatible trace elements and has time-integrated low Rb/Sr, Nd/Sm compared to the source
of intraplate oceanic islands (e.g., Hofmann, 1997). However, several studies (e.g., Ito
and Mahoney, 2005b; Meibom and Anderson, 2004) have argued that preferential melting
42
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
of incompatible element enriched, relatively fertile mantle lithologies at small degrees of
mantle melting could be at least partly responsible for the trace element and isotopic
differences between MORB and oceanic intraplate basalts (OIB). If this is the case, then
inversion of basalt compositions to infer mantle composition and melting processes will
not be straightforward (e.g., Stracke and Bourdon, 2009). Evidence for this kind of melting
behaviour is seen in the 143 Nd/144 Nd isotope compositions of clinopyroxenes from abyssal
peridotites, which extend to more radiogenic values than MORB from the same ridge
segment, suggesting that more enriched components are preferentially sampled during
melting and that the Nd isotope composition of the mantle may be more radiogenic than
that inferred from analysis of MORB alone (Salters and Dick, 2002; Snow et al., 1994;
Stracke et al., 2011; Warren et al., 2009).
At mid-ocean ridges, relatively high degrees of melting of a large volume of mantle, together with magma mixing during melt migration within the mantle and crustal magma
chambers have the effect of homogenising the magma compositions that are erupted at
the surface (e.g., Bryan, 1983; Dungan and Rhodes, 1978; Rhodes et al., 1979; Rubin
and Sinton, 2007; Sinton and Detrick, 1992; Walker et al., 1979). As a result, MORB are
unlikely to preserve an accurate picture of the degree of heterogeneity of the upper mantle
beneath spreading ridges. In contrast, magmas erupted on small off-axis seamounts apparently result from melting of smaller volumes of mantle. Most seamounts form within
5-15 km of the ridge axis (Scheirer and Macdonald, 1995), whereas the width of the melt
lens beneath the EPR is on average about 0.5 km, meaning that most seamount magmas
by-pass the main axial magma chamber system beneath spreading ridges. Seamount lavas
therefore undergo less mixing during melting and melt migration, and should more faithfully record the heterogeneity of the upper mantle. Previous studies have shown that the
lavas erupted on seamounts on the flanks of Pacific spreading ridges have far more variable trace element and isotope compositions than MORB erupted at the adjacent ridge
(Batiza and Vanko, 1984; Graham et al., 1988; Niu and Batiza, 1997; Niu et al., 2002;
Zindler et al., 1984). Seamount lavas therefore offer an opportunity to study the degree
of mantle heterogeneity in much greater detail.
Most previous studies have analysed only a few samples from individual seamounts. In
order to determine more precisely the compositional contrast between mantle heterogeneities, their length scale and origin, and how these are sampled during mantle melting,
more detailed geochemical studies of individual seamounts are required. Here, we present
new major and trace element and Sr, Nd and Pb isotope data for a suite of lavas from
Seamount 6, a small seamount located 140 km east of the East Pacific Rise at about 13◦ N.
The range of lava compositions on this single seamount is comparable to that observed
in the NE Pacific seamount lava dataset. Our new data provide insights into the nature
Dissertation P.A. Brandl
43
4. Volcanism on the flanks of the East Pacific Rise
and scale of upper mantle heterogeneity, and its influence on mantle melting processes.
4.2 Geological setting
Seamount 6 is located in the eastern Pacific on the Cocos Plate at 12◦ 45’N, 102◦ 35’W
(Fig. 4.1). Although the seamount has recently been renamed ‘Baja A’ (see seamount
catalogue; http://earthref.org/SBN/), we use the name Seamount 6 in this manuscript to
ensure consistency with earlier publications. Seamount 6 is situated some 140 km east of
the East Pacific Rise (EPR) on a plate segment bordered by the Orozco Fracture Zone
to the north and Clipperton Fracture Zone to the south (see Fig. 4.1). The full spreading
rate along this segment of the EPR is about 10-11 cm a−1 , but spreading is asymmetric
with a spreading rate of 6.5 cm a−1 to the west and 4.5 cm a−1 to the east (Choukroune
et al., 1984).
Seamount 6 consists of three coalesced volcanic edifices, termed 6W, 6C, and 6E (from
west to east) that are aligned parallel to the relative motion of the Cocos Plate. The
largest of the three cones is 6C with a basal diameter of 9.6 km, a volume of about
52 km3 and an elevation of ∼1,300 m above the seafloor (Batiza et al., 1989; Batiza and
Vanko, 1984). The eastern and western edifices are significantly smaller with volumes of
about 21 km3 and 22 km3 , and elevations of 750 m and 420 m above the seafloor, respectively (Batiza et al., 1989). Cone 6C has a prominent rift zone on its north side that
is oriented subparallel to the trend of the East Pacific Rise (340◦ ; Batiza et al., 1989).
Lavas from Seamount 6 were dredged during cruise RISE III Leg 3 of R/V New Horizon
in 1979 and also during Leg 3 of the CERES expedition in 1982. A detailed photographic
and submersible study (with submarine Alvin) followed during cruise AII-112-18 with
R/V Atlantis II in 1984, and cruise A132-17 in 1995. As a result of submersible observations and analysis of samples collected during these cruises, the geological structure and
petrological evolution of Seamount 6 is relatively well known (e.g., Fig. 4.2; Aggrey et al.,
1988; Batiza, 1980; Batiza et al., 1989; Batiza and Vanko, 1984; Graham et al., 1987,
1988; Honda et al., 1987; Zindler et al., 1984).
Seamount 6 is situated on oceanic crust that formed approximately 3.0 Ma ago during
magnetic anomaly 2’, which represents an upper limit on the age of the oldest lavas.
The seamount is composed entirely of normally polarised lavas, and is partly built on
negatively magnetised seafloor, indicating that the seamount was not formed directly
at the EPR axis. The magnetic data indicate that Seamount 6 formed within magnetic
anomaly 2’, no further than about 45 km from the EPR axis, on seafloor that was less
than about 1 Ma old (McNutt and Batiza, 1981; McNutt, 1986). An 40 Ar/39 Ar analysis
44
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.1: Bathymetric map of the Nazca Plate showing Seamount 6 and surrounding seafloor between
the Clipperton Fracture Zone and the Orozco transform fault. The samples analysed in this study are
all from the southern summit region of volcano 6C. Map prepared using GMT (Wessel and Smith, 1991,
1998).
of a sample of alkali basalt (19-23) from Seamount 6 yielded a ‘model age’ of <2 Ma
(Honda et al., 1987). The same sample was dated at 500±500 ka using the 3 He/4 He disequilibrium dating method (Graham et al., 1987). Other 3 He/4 He disequilibrium ages for
alkalic lavas from Seamount 6 range from 3 to 900 ka (Graham et al., 1987). Based on
the thickness of Fe-Mn coatings and degree of sediment cover, the three volcanic edifices of
Seamount 6 are approximately the same age, and lavas from the summit region of Seamount
6 are younger than those from the lower flanks (Batiza et al., 1989).
On Seamount 6, a relationship between lava composition, morphology and stratigraphy
has been documented by combined submersible, geophysical and geochemical studies
(Fig. 4.2a; Batiza et al., 1989; Batiza and Vanko, 1984, 1983; Graham et al., 1987,
1988; Honda et al., 1987; Maicher et al., 2000; McNutt, 1986; Zindler et al., 1984). In-
Dissertation P.A. Brandl
45
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.2: (a) Schematic E–W cross-section through Seamount 6 (not to scale) after Batiza and Watts
(1986). (b, c) Variation in primitive mantle normalised La/Sm and K/Ti ratios of Seamount 6 lavas
with bathymetric depth. Based on previous analyses of lavas from Seamount 6, the main edifice was
considered to consist of tholeiitic N-MORB and T-MORB with (La/Sm)N <1.8 and K/Ti <0.5, and only
alkalic E-MORB above 2000 m water depth. Our new data (diamond symbols) show that tholeiitic and
transitional lavas also occur above 2000 m. Data for Seamount 6 lavas from Batiza et al. (1989), Batiza
and Vanko (1984), Zindler et al. (1984).
tensely weathered pillow lavas composed of incompatible trace element depleted tholeiites
(N-MORB) with thick Fe-Mn coatings make up the lower slopes of 6C (below about
2,300 m water depth), and probably also much of the main edifices of 6E and 6W. The
summit regions of 6C and 6E are built of visibly younger hyaloclastites, sheet flows, pahoehoe lava and lobate pillows, and on 6C most of these consist of alkalic lavas of enriched
MORB (E-MORB) type. Previous geochemical studies of Seamount 6 lavas (e.g., Batiza
et al., 1989; Batiza and Vanko, 1984; Zindler et al., 1984) have shown that their chemical
and isotopic variability is far greater than that of MORB erupted at the adjacent northern EPR axis. Many Seamount 6 lavas have higher concentrations of incompatible trace
elements and higher 87 Sr/86 Sr and lower 143 Nd/144 Nd than most East Pacific Rise MORB.
In this study, we analysed 17 fresh volcanic glass samples that extend the known compositional range of Seamount 6 lavas in order to better constrain the melting dynamics in
this seamount.
4.3 Samples and methods
The samples analysed in this study were collected on dives 3009 to 3017 of submarine
Alvin during research cruise A132-17 of R/V Atlantis II in 1995. Most of these dives were
46
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
located at the upper, younger section of the southern slope of Seamount 6 (Fig. 4.1). The
terraced summit of Seamount 6 is predominantly characterised by pillow lavas, but also
hyaloclastites and sheet flows, which cap a previously flat-topped cratered edifice (Batiza
et al., 1989). Ferromanganese oxides have formed crusts up to 2 cm on lava and hyaloclastites (Table 4.1). There are no manganese nodules or ooze-mantling pavements and
the overall sediment cover in the study area is about 30-60%. In this study, we analysed
exclusively fresh volcanic glasses. A summary of lava flow type and petrography of our
new samples can be found in Table 4.1.
All major and trace element and Sr-Nd-Pb isotope analyses were carried out on fresh
glass fragments which were handpicked to avoid any with visible alteration. Major elements were analysed on single glass chips using a JEOL JXA-8200 Superprobe electron microprobe at the GeoZentrum Nordbayern in Erlangen. An acceleration voltage of
15 kV, a beam current of 15 nA, and a defocused beam (10 µm) were used. Counting times
were set to 20 s for peaks and 10 s for backgrounds, except for F and Cl for which counting times were increased to 40 and 20 s, respectively. Natural volcanic glass standards
(basaltic glass standard VG-A99 and rhyolitic glass standard VG-568), together with
mineral standards scapolite R-6600 (Smithsonian Institution) and apatite, chalcopyrite,
fluorite, rhodonite (P and H Developments) were used for calibration. Glass standards
VG-2, VG-A99 and VG-568 were analysed periodically as unknowns in order to monitor
the accuracy of the microprobe results (Table A7, Appendix). Major element data in Table
A7 of the Appendix represent averages of at least ten individual spot analyses per sample.
Trace element concentrations of glasses were determined using an Agilent 7500cs Quadrupole ICP-MS at the Institut für Geowissenschaften, Universität Kiel. Samples were digested following the pressurised HF-HClO4 -aqua regia acid procedure described by GarbeSchönberg (1993). Results of trace element analyses are given in Table A7 (Appendix);
precision and accuracy was checked by repeated analyses (n=5) of BHVO-2 (Govindaraju,
1995) and is better than 3 % (2SD) for most elements, and <7 % (2SD) for Cr, Cu, Sr,
Tl, V, W, Zn, and Zr.
The Sr, Nd, and triple-spike Pb isotope analyses were performed at the Max Planck
Institute for Chemistry in Mainz. Between 50 and 120 mg of sample material was cleaned
and ultrasonicated in ultrapure H2 O, and then leached in 2N HCl in an ultrasonic bath for
15 minutes. Samples were dissolved using distilled acids, and lead was separated first using
procedures described by Abouchami et al. (2000). Strontium and the rare-earth elements
were separated from the same sample dissolution by cation-exchange using AG50W-X8
resin and eluents of 2N HCl followed by 6N HCl. Neodymium was separated from the REE
fraction by cation exchange using 0.15N α-hydroxyisobutyric acid (α-HIBA) buffered
Dissertation P.A. Brandl
47
4. Volcanism on the flanks of the East Pacific Rise
at pH=4.5 as eluent. Procedural blanks were generally better than 165 pg for Sr and
2.62 pg for Nd.
Strontium, Nd, and Pb isotope compositions were measured on a ThermoFisher
TRITON TIMS operating in static multicollection mode. Sr and Nd were mass-bias corrected relative to 86 Sr/88 Sr = 0.1194 and 146 Nd/144 Nd = 0.7219. Standard runs of NIST
SRM-987 gave 0.710271±10 (2SD, n=14) and La Jolla Nd Standard runs yielded a value
of 0.511847±10 (2SD, n=12).
High-precision Pb isotope analyses were obtained using a 204 Pb-206 Pb-207 Pb triple spike
technique (Galer, 1999; Galer and Abouchami, 1998). Unspiked and spiked sample aliquots
were loaded onto Re single filaments along with silicagel-H3 PO4 activator. Pb blanks during this study were between 21 and 49 pg and are therefore negligible. Repeat measurements of the NIST SRM-981 Pb standard (n=24) during the period of analyses yielded
206
Pb/204 Pb, 207 Pb/204 Pb and 208 Pb/204 Pb of 16.9439±9, 15.5021±9, and 36.7328±24
(2SD), respectively, in agreement with values reported by Galer and Abouchami (1998).
In the following diagrams, we have included published isotope data for Seamount 6 lavas
(Graham et al., 1988; Zindler et al., 1984), re-normalised to our standard values where
necessary.
4.4 Results
4.4.1 Petrography
The petrography and phenocryst compositions of Seamount 6 lavas have been described
in previous studies (Batiza et al., 1989; Batiza and Vanko, 1984; Maicher et al., 2000).
Phenocrysts make up less than 10% of the volume of most lavas, and the most common
phenocryst assemblages in Seamount 6 lavas are plagioclase-olivine-spinel, plagioclaseclinopyroxene, and plagioclase-olivine-clinopyroxene (Batiza et al., 1989). All samples
analysed in this study were glassy pillow-rim fragments. A description of the petrography
and degree of alteration of the samples analysed is included in Table 4.1.
4.4.2 Major and trace element composition
Lavas from Seamount 6 range from subalkaline basalts to trachybasalts and basaltic trachyandesites on a volatile-free total alkali–silica (TAS) diagram (Fig. 4.3a; Le Maitre
et al., 1989). In our sample set, MgO varies between 4.6 and 9.8 wt. %; some samples
analysed by Batiza et al. (1989) extend to 3.1 wt. % MgO. TiO2 , Na2 O, and K2 O increase
(Fig. 4.3c, g and h), and CaO (Fig. 4.3f) decreases with decreasing MgO, whereas SiO2 ,
48
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Table 4.1: Lava flow type, alteration and petrography of Seamount 6 samples.
Sample
3009-1511
3009-1557
3010-1032
3010-1053
3010-1219
3010-1249
3010-1453
3010-1502
3011-1333
3013-1315
3013-1425
3014-1136
3015-1048
3015-1344
3015-1432
3015-1437
3016-1502
Lava type
‘Cowheart’ lava
Sheet flow lava
Sheet flow lava
Sheet flow lava
Sheet flow lava toe
Folded sheet flow lava
Sheet flow lava
Lobate flow lava
Folded sheet flow
Thin sheet flow
Huge lava ‘hollow bowl’
Sheet flow lava
Lava flow
Lava flow
Sheet flow lava
Lava flow
Sheet flow lava
Alteration
Mn-crust
Palagonite
2–3 mm
<1 mm
2–3 mm
1–2 mm
None
Some coating
<1 mm
1–2 mm
None
Some coating
Up to 5 mm
<1 mm
3–4 mm
1–2 mm
None
1–2 mm
Up to 5 mm
<1 mm
Up to 5 mm
<1 mm
Up to 5 mm
<1 mm
None
<1 mm
None
1–2 mm
None
<1 mm
2–3 mm
<1 mm
2–3 mm
1–2 mm
None
Some coating
Petrography
Phyric/aphyric
Minerals
Aphyric
Aphyric
Aphyric
(Aphyric)
Few Ol phenocr.
Aphyric
Aphyric
2–3% phyric
Plg+Ol
∼5% phyric
Plg(+Ol)
1–2% phyric
Plg+Ol
Aphyric
Aphyric
2–3% phyric
Plg
Aphyric
Aphyric
1–2% phyric
Plg
2–3% phyric
Plg
Aphyric
Al2 O3 , and FeOT (Fig. 4.3b, d and e) do not vary systematically with MgO. Compared to
most N-MORB from the EPR, lavas from Seamount 6 have higher Al2 O3 , Na2 O, K2 O and
K2 O/TiO2 , and lower FeOT for a given MgO. Major element compositions of Seamount 6
lavas generally overlap with those of other EPR near-ridge seamount lavas, although five
of the alkalic lavas (trachybasalts and basaltic trachyandesites) we have analysed have
higher K2 O and K2 O/TiO2 than even the most ‘enriched’ seamount lavas reported by
Niu and Batiza (1997); other Seamount 6 samples analysed by Batiza and Vanko (1984)
have K2 O up to 2.56 wt. %. The alkalic lavas (highest K2 O) also have the most evolved
compositions (lowest MgO contents; Fig. 4.3h). The samples generally do not define clear
arrays in major element diagrams, which suggests that they cannot be simply related by
fractional crystallisation from a common parental magma composition. The large range in
K2 O, Na2 O and K2 O/TiO2 is in any case difficult to explain by fractional crystallisation
processes, as discussed in more detail in section 4.5.1.
The concentrations and ratios of highly incompatible elements are far more variable in
Seamount 6 lavas than in MORB from the adjacent EPR. For example, Nb concentrations
in Seamount 6 lavas vary between 1.66 and 63.4 ppm, extending to higher concentrations
than those of the seamount lavas analysed by Niu and Batiza (1997), and previous analyses
of Seamount 6 lavas (Fig. 4.4, Fig. 4.5a,b). The samples with the lowest Nb concentrations are similar to those found in MORB from the EPR however, none of the Seamount 6
lavas have Nb concentrations as low (0.3 ppm) as some of the seamount lavas analysed by
Niu and Batiza (1997). The concentrations of incompatible elements are correlated with
major element compositions; samples with the highest Nb concentrations have the lowest
MgO, CaO and highest K2 O.
Dissertation P.A. Brandl
49
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.3: Major element compositions of Seamount 6 lavas, compared with other EPR seamount lavas
(grey circles; Niu and Batiza, 1997) and EPR axial lava (EPR FC-Suite: basalts and andesites of Regelous
et al., 1999; EPR N- to E-MORB: Niu et al., 1999 and Waters et al., 2011). Compared to lavas from
the EPR spreading axis and most other Pacific seamount lavas, Seamount 6 lavas have higher Na2 O and
K2 O (extending to basaltic trachyandesites in (a)), higher Al2 O3 and lower FeO.
Ratios of highly-to-moderately incompatible elements, for example La/Yb, Sm/Yb and
Nb/Zr ratios display negative correlations with MgO and CaO (Fig. 5c), and positive
correlations with incompatible element concentrations. Nb/Zr ratios vary by almost an
order of magnitude (0.025 to 0.20). To our knowledge, the range in trace element compo-
50
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.4: Trace element concentrations of Seamount 6 lavas, normalised to primitive mantle
(McDonough and Sun, 1995). Fields for lavas from other Pacific near-ridge seamounts (Niu and Batiza,
1997), and lavas from the nearby EPR axis (Niu et al., 1999; Regelous et al., 1999; Waters et al., 2011)
shown for comparison. Compared to the latter, Seamount 6 lavas display a much greater range in trace
element composition. Alkalic lavas from Seamount 6 extend to higher Nb, Th, Rb and La/Yb than other
Pacific seamount lavas. The most incompatible element depleted tholeiites from Seamount 6 have similar
trace element compositions to tholeiites erupted at the EPR axis; highly depleted lavas with BaN 0.2–1.0,
such as are found on some other EPR seamounts, are apparently not present on Seamount 6.
sitions in Seamount 6 lavas is greater than that reported for any other known near-ridge
NE Pacific seamount (Niu et al., 2002; Niu and Batiza, 1997). Seamount 6 lavas extend
to highly enriched compositions, even though the upper mantle close to spreading ridges
far from hotspots is often assumed to be relatively homogeneous and depleted (Hertogen
et al., 1980; Saunders et al., 1988; Schilling et al., 1983; Zindler et al., 1984). The Ce/Pb
and Nb/U ratios of our samples (22-29 and 36-46, respectively) lie within the range defined by most fresh oceanic basalts (25±5 and 45±10 for Ce/Pb and Nb/U respectively;
Hofmann, 1997).
We have divided our samples into tholeiitic basalts, transitional basalts and alkalic lavas
(trachybasalts and basaltic trachyandesites), on the basis of their SiO2 and total alkali
contents (Fig. 4.3a). These three groups correspond in their chondrite-normalised La/Sm
ratio to N-MORB (La/Sm)N <0.7; tholeiitic basalts, Fig. 4.2b), T-MORB (0.7 to 1.8;
transitional basalts), and E-MORB (>1.8; alkalic basalts and differentiates), as defined
by Schilling et al. (1983). Although previous studies of Seamount 6 lavas found exclusively
E-MORB at water depths of <2,000 m, our new data show that lavas from the summit
regions also include N-MORB and T-MORB (Fig. 4.2). The entire Seamount 6 database
includes samples from the summit region and the deepest flanks, and so is likely to be
Dissertation P.A. Brandl
51
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.5: (a) and (b) Zr and Nb concentrations in basalts and andesites from the EPR axis
at 10◦ 30’N (‘FC-Suite’) both vary by a factor of about 6, and show little variation in Nb/Zr (c), which
can be explained by fractional crystallisation (Regelous et al., 1999). In contrast, concentrations of highly
incompatible elements Nb and Th in Seamount 6 lavas vary by almost two orders of magnitude, and the
wide range in Nb/Zr is inconsistent with fractional crystallisation of observed phenocryst phases (see text
for discussion). Alkalic lavas from Seamount 6 have the highest Nb concentrations, the highest Nb/Zr,
La/Sm and K/Ti.
representative of much of the range in composition of the exposed lavas.
4.4.3 Radiogenic isotopes
The Sr, Nd and Pb isotope ratios of the Seamount 6 lavas analysed in this study cover
most of the range defined by previous isotope analyses of lavas from this seamount (Fig.
4.6). However, our new isotope data display less scatter in Pb-Pb, Sr-Pb and Nd-Pb
isotope spaces compared to existing data, which is likely due to the higher precision of
our Pb-triple spike analyses compared to older conventional Pb isotope data, together
with the fact that some previous isotope analyses of Seamount 6 lavas were carried out on
variably altered whole-rock samples (Zindler et al., 1984). The tholeiitic and transitional
basalts have Sr, Nd and Pb isotope compositions that overlap those of MORB erupted
at the adjacent EPR axis, but the alkalic lavas extend to higher 87 Sr/86 Sr, 206 Pb/204 Pb,
207
Pb/204 Pb and 208 Pb/204 Pb, and lower 143 Nd/144 Nd (Fig. 4.6). The combined Seamount
6 dataset displays almost the entire range of Sr, Nd and Pb isotope compositions observed
in all other near-ridge NE Pacific seamounts (Fig. 4.6).
52
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.6: Sr, Nd and Pb isotope composition of lavas from Seamount 6 (Allan et al., 1989;
Batiza and Vanko, 1984; Graham et al., 1987; Niu et al., 2002; Zindler et al., 1984; this study) display
a larger range than MORB from the adjacent EPR axis, which is comparable to the range in the entire
EPR seamount database. Alkalic lavas from Seamount 6 have higher 87 Sr/86 Sr, lower 143 Nd/144 Nd and
more radiogenic Pb isotope compositions than transitional and tholeiitic lavas. The alkalic lavas also have
the highest Nb concentrations, highest Nb/Zr, and lowest CaO and MgO. Regression lines through the
Pb isotope data are calculated using the Williamson/Minster method (Table 4.2).
The Seamount 6 samples display remarkable correlations between Sr, Nd, Pb isotope
ratios, major element concentrations and the concentrations and ratios of incompatible
elements. Alkalic lavas with higher concentrations of the incompatible elements such as
Nb have higher Nb/Zr, La/Sm, higher 87 Sr/86 Sr (Fig. 4.6c), lower 143 Nd/144 Nd (Fig. 4.6a)
and more radiogenic Pb isotope ratios (Fig. 4.6b, d). These lavas also have the highest
K2 O and lowest MgO and CaO contents (Fig. 4.3f,h). Similar correlations are seen in the
NE Pacific near-ridge seamounts studied by (Niu and Batiza, 1997; Niu et al., 2002), but
the correlations in our dataset are far better defined. The significance of the observations
above are discussed below in detail.
4.5 Discussion
4.5.1 The effects of fractional crystallisation
The low Mg# values of the Seamount 6 lavas (71 to 50) indicate that none are likely to
represent primary melts in equilibrium with mantle olivine (Fo = 89), and all have probably undergone some degree of crystal fractionation before eruption. However, in major
Dissertation P.A. Brandl
53
4. Volcanism on the flanks of the East Pacific Rise
element diagrams our samples do not form clearly defined arrays, and are displaced from
the fields defined by basalts, basaltic andesites and andesites from the nearby EPR axis
at 10◦ 30’N (Fig. 4.3 ‘EPR FC-Suite’; Regelous et al., 1999). Compared to these samples,
the Seamount 6 lavas have lower FeOT , slightly lower SiO2 and CaO, higher Al2 O3 and
Na2 O, and far higher K2 O for a given MgO (Fig. 4.3b and d-h).
The major and trace element variations within Seamount 6 lavas are difficult to explain by
fractional crystallisation. For example, to generate the extreme range in Nb concentrations
(1.66 to 63.4 ppm; Fig. 4.5a) would require approximately 97 % crystallisation, even assuming Nb to behave as a perfectly incompatible element. In the Nb-Zr diagram (Fig. 4.5b
‘EPR FC-Suite‘), the basaltic to andesitic lavas from the EPR axis at 10◦ 30’N (Regelous
et al., 1999) define an array passing through the origin, consistent with fractional crystallisation as the major control on Nb and Zr concentrations. In contrast, Seamount 6 lavas
have a far greater range in Nb contents, which correlate with the Nb/Zr ratio. The large
range in incompatible trace element ratios such as Nb/Zr and La/Yb cannot be explained
by fractionation of the likely crystallising phases, olivine, plagioclase, clinopyroxene and
spinel, in which these elements are highly incompatible.
As noted above, the isotope compositions of Sr, Nd and Pb are correlated with incompatible trace element ratios and major and trace element concentrations, and these correlations are inconsistent with closed system crystal fractionation. Although assimilation of
seawater-altered oceanic crust could result in higher 87 Sr/86 Sr in the more evolved lavas,
crustal contamination is unlikely to significantly affect Nd isotope compositions, nor the
ratios of incompatible, immobile trace elements such as Nb/Zr (Bienvenu et al., 1990).
We conclude that processes other than fractional crystallisation and assimilation must
be responsible for much of the range in composition of most Seamount 6 lavas. However,
three alkalic lava samples (3015-1344, -1432, -1437) have higher MgO for a given K2 O
and CaO compared to other Seamount 6 lavas. These three samples also have higher concentrations of Ni, lower concentrations of incompatible elements (e.g., Ba and Zr), but
similar incompatible trace element and isotope ratios to other Seamount 6 alkalic lavas,
and these samples may therefore represent a group of less fractionated alkali basalts.
In summary, although some of the major element variation in Seamount 6 lavas may
be the result of crystal fractionation, the correlations of MgO and CaO with concentrations of highly incompatible elements such as K and Nb and trace element (Fig. 4.5c) and
isotope ratios, suggest that much of the major element variation in Seamount 6 lavas is
not the result of fractional crystallisation, but must be the result of other processes. In the
following section, we examine these other processes using incompatible trace element and
isotope ratios which are unaffected by fractional crystallisation; but we note here that if
54
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
these processes commonly occur in oceanic basalt suites, then the fractional crystallisation
history inferred from major element data alone may not always be accurate.
4.5.2 The role of mixing processes
In the Seamount 6 dataset, incompatible trace element and isotope ratios are correlated
with the concentrations of major elements. The good correlations between MgO, CaO and
Nb, Nb/Zr and 87 Sr/86 Sr are more easily explained if a more fractionated (lower MgO, and
CaO; Fig. 4.5c, 4.7b) melt with higher incompatible element concentrations (Fig. 4.7c),
higher Nb/Zr and 87 Sr/86 Sr (Fig. 4.6c) mixed with a less fractionated, relatively depleted
melt. The mixing must have occurred shortly after fractionation, and thus likely at high
levels in the crust, in order to preserve the correlations with major element contents. This
type of mixing, in which two distinct endmember melts or mantle source compositions
are mixed together in variable proportions, will result in straight mixing lines in elementelement diagrams, and in all ratio-ratio diagrams in which the denominator element is
the same (Langmuir et al., 1978). For Seamount 6 lavas, two-component mixing appears
to explain most of the combined major and trace element and isotope variations (e.g.,
Fig. 4.7a: La/Sm vs. 1/Sm). Previous studies of Pacific near-ridge seamount lavas (e.g.,
Batiza and Vanko, 1984; Niu and Batiza, 1997; Zindler et al., 1984) have also argued that
much of the compositional variation can be explained by mixing depleted and enriched
endmember compositions.
Assuming that the variations in Seamount 6 lavas are indeed the result of mixing of two
endmember melt compositions, we can use the curvature and orientation of these mixing
arrays to estimate the approximate compositions of the mixing endmembers (e.g., Nauret et al., 2006). For example, in Figure 4.7e, the 206 Pb/204 Pb composition of the high
206
Pb/204 Pb endmember must lie in the range between the highest 206 Pb/204 Pb ratio of the
analysed samples (18.95) and the intersection of an extension of the Yb/Pb–206 Pb/204 Pb
array with the y-axis at Yb/Pb = 0 (i.e. 19.03). Similarly, the 206 Pb/204 Pb of the more
‘depleted’ endmember is constrained to lie in the range 18.2–18.0 from Figure 4.7d.
Additional information on the compositions of the mixing endmembers can be gained
from the orientation of the mixing arrays in diagrams such as Zr/Rb - Nb/Sm (Niu and
Batiza, 1997), where the hyperbolic arrays constrain the Zr/Rb and Nb/Sm values of the
enriched and depleted components (<20 and <0.5, respectively; see Fig. 4.7f). Similarly,
the curvature of the array requires that the value of (SmE x RbD )/(SmD x RbE ) is less
than 0.05, where subscripts D and E refer to the concentration of an element in the depleted (high Zr/Rb) and enriched (high Nb/Sm) endmembers, respectively.
Dissertation P.A. Brandl
55
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.7: Calculating the chemical composition of mixing endmembers. The systematic variations in major and trace element and isotope composition in Seamount 6 lavas are most easily explained
by two-component mixing. In trace element ratio diagrams in which the same element is used as denominator (a,d,e), Seamount 6 lavas define approximately linear arrays, consistent with such mixing. The
intercept of these arrays with the y axis in (d) and (e) can be used to estimate the Pb isotope composition
of the mixing endmembers, and shows that these cannot be very different from the Pb isotope compositions of lavas with the most- and least-radiogenic Pb. Hyperbolic mixing arrays in diagrams such as 4.6f
are also consistent with mixing, and constrain the trace element composition of the mixing endmembers
(see text). Correlations of incompatible trace element and isotope ratios with major element concentrations (e.g., 4.6f) show that mixing likely took place at high levels in the crust, after most fractional
crystallisation had occurred. Data for EPR seamounts from Niu and Batiza (1997) and Niu et al. (2002);
data for EPR axis lavas from Castillo et al. (2000), Niu et al. (1999), Regelous et al. (1999) and Waters
et al. (2011).
In Pb isotope diagrams (Fig. 4.6b,d), the Seamount 6 lavas define highly linear arrays
(see Table 4.2). Linear arrays defined by oceanic basalts in 207 Pb/204 Pb–206 Pb/204 Pb diagrams have sometimes been interpreted as isochrons (Chase, 1981; Gale and Mussett,
1973; Gast et al., 1964; Tatsumoto, 1978). For Seamount 6 lavas, the model source age calculated from the slope of the data in Figure 4.6b is 2530±29 Ma, but the slope of the data
in the 208 Pb/204 Pb - 206 Pb/204 Pb diagram (Fig. 4.6d) then implies a source 232 Th/238 U
56
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
ratio (κ) of ∼4.04 (Table 4.2). This is higher than the measured κ values in most Seamount
6 lavas, which represent an upper bound on the κ value of their mantle source due to the
higher incompatibility of Th relative to U during melting. The unrealistically high κ inferred for the source of Seamount 6 lavas, together with the evidence for mixing discussed
above, strongly suggests that the Pb isotope arrays represent mixing lines and convey no
source age information (e.g., White and Schilling, 1978; Abouchami et al., 2000).
Table 4.2: Linear least-squares regression parameters for Pb isotope data‡ .
n
14
‡
Slope in 206 Pb/204 Pb vs.
207
Pb/204 Pb
0.1671±0.0029
χ2r
τ (Ga)
0.51
2.530
±0.029
Slope in 206 Pb/204 Pb vs.
208
Pb/204 Pb
1.1542±0.0098
χ2r
κ
κ*
1.1886
4.160
±0.042
3.23
±0.27
Regression parameters calculated using the Williamson/Minster method (Williamson, 1968). n: number
of samples; χ2r : reduced chi-squared (MSWD); τ : model age; κ*: atomic
232
Th/238 U ratio derived from
measured Th/U (ppm).
4.5.3 Highly heterogeneous mantle beneath Seamount 6
The inferred difference in composition of the endmember melt compositions which were
mixed to form the arrays in Figures 4.5, 4.6 and 4.7, including the order of magnitude difference in Nb/Zr ratio (Fig. 4.5) and large difference in Pb isotope composition
(Fig. 4.7d,e), indicate that much of the trace element and isotopic heterogeneity observed
in Seamount 6 lavas is ultimately the result of heterogeneity in the mantle source of the
lavas. The estimates of the Pb isotope compositions of the mixing endmembers (Section
4.5.2) suggest that Seamount 6 lavas preserve most of the original Pb isotope variation
in the melts that were mixed during the most recent mixing event that gave rise to the
observed correlations between major and trace element and isotope composition. The ‘enriched’ endmember melt composition is more enriched (e.g., lower 143 Nd/144 Nd, higher
Nb/Sm) than the most enriched E-MORB erupted at the adjacent EPR axis, and the
depleted endmember is at least as depleted as the most depleted northern EPR N-MORB
(Fig. 4.6, 4.7). For Seamount 6 lavas, the correlations of isotope and incompatible trace
element ratios with major element compositions show that mixing took place at relatively
high levels in the crust, after most fractional crystallisation had occurred. The range in
primitive melt compositions produced during melting is therefore likely to have been far
greater, and the range in composition of the mantle source greater still. The full spectrum
of mantle heterogeneity is unlikely to be observed in the lavas erupted at the surface,
due to homogenisation effects during melting and melt transport (e.g., Rubin and Sinton,
2007; Rubin et al., 2009), even though these effects are less pronounced beneath off-axis
seamounts than beneath the adjacent spreading axis.
Dissertation P.A. Brandl
57
4. Volcanism on the flanks of the East Pacific Rise
Lavas from Seamount 6 span almost the entire range of compositions seen in the Niu
and Batiza (1997) dataset for lavas from near-ridge seamounts over a wide area of the
EPR flanks, which suggests that the length scale of mantle heterogeneity is smaller than
or comparable to the size of the melting region beneath a single seamount (less than a
few tens of km). The overlap in composition between lavas from Seamount 6, and from
other small seamounts on the flanks of the EPR, suggests that the entire northern EPR
area is underlain by similarly heterogeneous mantle.
The compositional range within the seamount lavas studied by Zindler et al. (1984), Niu
and Batiza (1997) and Niu et al. (2002) was ascribed to mixing of melts from a heterogeneous mantle source (‘melting-induced mixing’; Niu and Batiza, 1997). In detail however,
the range in composition in the entire northern EPR seamount dataset is inconsistent
with simple two-component mixing of melts from different mantle lithologies, or melting
of heterogeneous mantle in which enriched and depleted materials are mixed together
in variable proportions (‘source mixing’) but contribute equally to melting. For example,
EPR seamount lava compositions define curved arrays in diagrams such as La/Nd–Yb/Nd
(Fig. 4.8b–d), which are inconsistent with such mixing. Similar curved arrays have been
reported for lavas from the fossil Galapagos Rise spreading axis in the south-eastern Pacific. Haase et al. (2011b; see chapter 5) have shown that these can be produced by melting
of heterogeneous mantle if more enriched lithologies have lower solidus temperature and
contribute more to melting at small degrees of melting. In the following section, we therefore examine the effects of melting a ‘plum-pudding’ mantle in more detail, in order to
gain insights into the nature of upper mantle heterogeneity beneath the NE Pacific, the
length scale of heterogeneity, the chemical and isotopic variability, and the influence of
mantle heterogeneity on the melting process at spreading ridges. We consider here the
entire northern EPR seamount dataset (Graham et al., 1987; Zindler et al., 1984; Niu and
Batiza, 1997; Niu et al., 2002), because this includes more depleted lavas than are found
on Seamount 6, and is therefore more likely to be representative of the complete range of
melt compositions produced beneath the EPR.
4.5.4 Melting of a heterogeneous mantle
Melting models and input parameters
There is increasing evidence to suggest that the upper mantle source of oceanic basalts
consists of relatively fertile and more refractory lithologies which have different solidus
temperatures and thus begin melting at different depths during adiabatic upwelling. For
example, lavas that are apparently produced by small degrees of mantle melting tend
to have more enriched incompatible trace element and isotope compositions than those
58
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
resulting from larger average degrees of mantle melting (Haase et al., 2011b; see chapter 5). Conversely, lavas produced by melting of previously melted mantle tend to have
relatively depleted compositions (e.g., Wendt et al., 1999). Similarly, abyssal peridotites,
which are thought to represent the residues of melting at spreading ridges, apparently extend to higher 143 Nd/144 Nd and less radiogenic Os isotope compositions than associated
MORB (Liu et al., 2008; Salters and Dick, 2002). These observations can be explained if
incompatible trace element enriched lithologies with low 143 Nd/144 Nd have lower solidus
temperatures, so that they contribute more to melting than relatively depleted, less fertile
lithologies which are concentrated in the melting residues.
Table 4.3: Input parameters used in the melting models illustrated in Figure 4.8∗ .
Mass porosity
Input compositions
Case 4.8b
Mass fraction
Case 4.8c (I)
Case 4.8c (II)
Case 4.8c (III)
Onset of melting
Case 4.8d (I)
Case 4.8d (II)
Melt productivity
End of melting
∗
Spinel-peridotite
0.1%
Depleted DMM (Workman and Hart,
2005)
Depleted DMM (Workman and Hart,
2005) depleted by 5% batch melting
93%
95%
90%
50%
2.5 GPa
2.5 GPa
2.5 GPa
Power-law fashion 0.2-2.0% km−1 (Asimow et al., 2004)
When F=0.2
Pyroxenite
0.1%
2.0 Ga old recycled MORB (Stracke
and Bourdon, 2009)
Global MORB average Arevalo Jr. and
McDonough (2010)
7%
5%
10%
50%
Prior to onset of peridotite
4.0 GPa
After onset of peridotite
Linearly from 0.45-2.12% km−1 (Pertermann and Hirschmann, 2003)
For full discussion and more detail on input and modelling parameters, the reader is referred to section
4.6 and Stracke and Bourdon (2009).
Several recent studies have attempted to quantify the range in melt compositions produced during melting of ‘plum pudding’ mantle (Ito and Mahoney, 2005a,b; Pearce, 2005;
Phipps Morgan, 2001; Phipps Morgan and Morgan, 1999; Stracke and Bourdon, 2009).
In all these models, important controls on the range of melt compositions are the trace
element and isotope compositions of the different lithologies, their mineralogy and melting
behaviour, the average degree of melting, and the extent to which melts are mixed together before being erupted at the surface. We have used the melting model of Stracke and
Bourdon (2009) to examine the range of melt compositions produced during melting of a
two-component mantle. In this model, the mantle source consists of two lithologies which
have different solidus temperatures and thus begin melting at different depths during
mantle upwelling. At a given depth, these lithologies will contribute unequally to melting.
The melt produced from both lithologies at all depths within the melting column are
pooled before eruption, and the ‘melt extraction trajectories’ (see Phipps Morgan, 2001)
Dissertation P.A. Brandl
59
4. Volcanism on the flanks of the East Pacific Rise
followed by melts produced by increasing degree of upwelling and melting of this heterogeneous mantle thus evolve from relatively enriched to more depleted compositions. In
section 4.5.4 we model the range of melt compositions produced during melting of a mixed
pyroxenite-peridotite mantle, using mineral and melting modes and partition coefficients
as defined by Stracke and Bourdon (2009) and reported in section 4.6. We investigate the
effects of varying melt column length, proportion of pyroxenite, difference in solidus temperature, and the trace element and isotope compositions of the endmember lithologies,
in order to try to place some constraints on the physical processes of melting and the
nature of mantle heterogeneity beneath Seamount 6 and the adjacent spreading ridge.
Modelling results and implications for seamount lava petrogenesis
The ‘melt extraction trajectories’ in Figure 4.8 illustrate the range of pooled melt compositions produced during progressive melting of a mixed peridotite-pyroxenite mantle.
Small degree melts formed deep within the melting column are dominated by the enriched pyroxenite component, and with increasing total degree of melting (and thus also
increasing melting column length), pooled melt compositions are increasingly dominated
by the more depleted component. An important result is that variable degrees of melting
of heterogeneous mantle can broadly reproduce the curved data array defined by EPR
seamount lavas in the La/Nd–Yb/Nd diagram (Fig. 4.8). As discussed previously, these
curved arrays cannot be produced by two component mixing of sources or melts. The
effects of increasing the difference in solidus temperature, or the proportion of pyroxenite
relative to peridotite, is to increase the curvature of these melt extraction trajectories
(Fig. 4.8c,d). As discussed in detail by Stracke and Bourdon (2009), if the contrast in
melting behaviour is too great, the pyroxenite component is completely exhausted before
peridotite melting begins, and clear correlations between trace element and isotope ratios, as observed for Seamount 6 lavas (Fig. 4.6c, 4.7c, 4.8f) will not be observed in the
pooled melts. These correlations also require the enriched component to be a volumetrically minor component of the mantle (Fig. 4.8c). Although we have used a pyroxenite
as the fertile lithology in the modelling, a fertile pyroxene- and garnet-rich peridotite
could also explain the observed range in lava compositions, provided that this peridotite
is enriched in incompatible elements and has a sufficiently lower solidus than the more
refractory matrix. Figure 4.8 shows that pooled melt compositions produced by melting of
a two-component mantle vary significantly with variations in the melting column length,
and thus likely with differences in the distance to the ridge axis (age of the lithosphere).
Beneath thicker lithosphere on the ridge flanks outside the neovolcanic zone, the melting
column is shorter, the average degree of melting is lower (The MELT Seismic Team, 1998),
and the melts produced by melting of two-component mantle are more enriched (higher
La/Nd, lower Yb/Nd; Fig. 4.8a).
60
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Figure 4.8: Results of a melting model (Stracke and Bourdon, 2009) in which mantle consisting
of two different lithologies having different solidus temperature and composition are progressively melted
during upwelling and decompression. Curves in panel (a) shows La/Nd and Yb/Nd compositions of
melts derived from mantle consisting of 7% ‘pyroxenite’ (green) and 93% peridotite (blue). Pyroxenite is
assumed to begin melting at greater depth than peridotite, so that with increasing melting column length
(decreasing lithosphere thickness), the composition of the pooled melt from both lithologies (red curve)
is increasingly dominated by the contribution from incompatible trace element depleted peridotite and so
evolves to lower La/Nd, higher Yb/Nd. Thus, unlike two-component mixing of melts or sources, melting
of a two-component mantle leads to curved arrays on this diagram, which approximate the data array
defined by Pacific near-ridge seamounts (b). For a greater compositional contrast between pyroxenite and
peridotite, shorter melting column lengths are required to produce a given range in La/Nd and Yb/Nd
(panel b). Curves in (c) and (d) show the effects of varying the proportion of pyroxenite relative to
peridotite, and changing the difference in solidus temperature of each lithology. If the proportion of the
pyroxenite is too high, or if the difference in melting behaviour of the two lithologies is too great (purple
curves in (c) and (d)), and a clear negative correlation between Yb/Nd and La/Nd is not obtained (see
Stracke and Bourdon (2009) for full discussion). Melting of two-component mantle can also explain the
hyperbolic array defined by all EPR seamount lavas in (e), and the correlations between trace element
and isotope ratios, such as those in (f). In contrast, the variations in Seamount 6 lavas are apparently
best explained by incomplete mixing of melts from various depths in the melting column (black dashed
lines in panel (a)). Incomplete mixing may also account for some of the variation within the entire EPR
seamount lava dataset, but among these samples, the effects of melting heterogeneous mantle is clearer,
because most other seamounts are represented by less than 3 samples. The mineralogy, melting mode
and trace element compositions of pyroxenite and peridotite are from Stracke and Bourdon (2009) and
can be found in section 4.6. Other parameters used in the melting model are listed in Table 4.3. Data for
Seamount 6 lavas from Batiza et al. (1989); Batiza and Vanko (1984); Zindler et al. (1984), and this study;
data for other EPR Pacific seamounts (Niu and Batiza, 1997; Niu et al., 2002) and the EPR (Regelous
et al., 1999) shown for comparison.
Dissertation P.A. Brandl
61
4. Volcanism on the flanks of the East Pacific Rise
To some extent, this effect may explain the range in lava compositions erupted on Pacific
near-ridge seamounts. For the particular melting model illustrated in Figure 4.8a, the
large variations in La/Nd observed in EPR seamount lavas require variations in melting
column length of more than 40 km. The lavas analysed by Niu and Batiza (1997) and Niu
et al. (2002) are from seamounts located on crust younger than 4 Ma, which represents
a maximum constraint on their age; some lavas may therefore have been erupted up to
200 km off-axis. However, the distribution of seamount density with crustal age (Scheirer
and Macdonald, 1995) indicates that the majority of EPR seamounts apparently formed
within 5–15 km of the spreading ridge axis (corresponding to crustal ages of 0.1–0.3 Ma).
Some seamounts must have formed on older, thicker lithosphere (Scheirer and Macdonald,
1995), although still within the 100–200 km wide zone on either side of the EPR within
which partial melt has been detected seismically (The MELT Seismic Team, 1998). Nevertheless, large differences in melting column length over horizontal distances of 5–15
km close to the axis are unexpected, and may indicate that the compositional contrast
between the peridotite and pyroxenite lithologies is much greater than assumed in our
model. Figure 4.8b shows the effects of increasing the difference in trace element composition between fertile and refractory lithologies - for a greater contrast, a smaller range in
melt column length is required to produce a given range in La/Nd.
Seamount 6 apparently formed on young (<1 Ma old) oceanic crust outside the neovolcanic zone, over a period of about 1 Ma. Little change in lithosphere thickness or melting
column length is expected over this short time period, yet Seamount 6 lavas display a wide
range in compositions. The variations within lavas from this single seamount may best be
explained by incomplete pooling (mixing) of melts from different depths within the melting column. We have argued above that the chemical and isotopic variation in Seamount
6 lavas is the result of mixing at crustal levels of melts from enriched and depleted mantle
lithologies, and that much of the original heterogeneity is preserved. The orientation of
the mixing array in the La/Nd–Yb/Nd diagram is qualitatively consistent with mixing
between melts of peridotite (lower La/Nd higher Yb/Nd) and pyroxenite (Fig. 4.8a,b). In
contrast, in the Niu and Batiza (1997), Niu et al. (2002) EPR seamount lava dataset, the
effects of magma mixing, which would tend to obscure the melt extraction curves, is less
clear because most of these different seamounts are represented by only 1-3 samples.
Sampling of melts at different depths within the melting column might also explain the
apparent geochemical evolution of Seamount 6. A possible tectonic model is that lithospheric weaknesses allow melt from the underlying partially molten mantle to drain to
the surface - first relatively depleted melts from melting of peridotite at shallow levels in
the mantle are extracted, at a greater rate than melt is replenished by mantle upwelling,
followed by more enriched melts dominated by the pyroxenite component from deeper in
62
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
the melting column. Once the available melt has been extracted, the seamount volcano becomes extinct. Removal of low-degree, incompatible trace element enriched partial melts
by seamount volcanism on the flanks of spreading ridges could significantly change the
average composition of the remaining melt, which is focussed towards the ridge axis and
erupted within the neovolcanic zone.
On Seamount 6, highly depleted lavas with Yb/Nd ratios of 0.6, similar to those on some
other EPR seamounts (Fig. 4.8), have not been found. We suggest that very depleted
lavas may only be erupted on seamounts that initially formed within the neovolcanic zone
of the adjacent spreading ridge, where the underlying melting column is longest. Mixing
arrays between melts derived from higher degrees of melting of enriched and depleted
materials are predicted to have a near-vertical orientation in the La/Nd–Yb/Nd diagram
(Fig. 4.8a). Additional detailed geochemical and geochronological studies of individual
seamounts are needed to test this idea.
As discussed earlier, the Seamount 6 lavas with the most radiogenic 87 Sr/86 Sr, highest
K/Ti and La/Sm also tend to have the lowest MgO and CaO. Similar correlations between major element compositions and incompatible trace element and isotope ratios
were observed for the seamount lavas studied by Niu et al. (2002), who suggested that
more enriched, volatile-rich melts cool to lower temperature and thus fractionate more.
Alternatively, primitive melts of more enriched mantle lithologies may have lower MgO
before fractional crystallisation at crustal levels, or they may undergo higher degrees of
fractionation during melt migration to the surface, since they are predicted to have been
produced at deeper levels in the melting column and thus have a longer distance to travel
to the surface.
4.5.5 Implications for the use of oceanic lavas as probes of mantle
composition
Our results show that Seamount 6 lava compositions can be explained by incomplete mixing of melts produced by melting of a two-component mantle, consisting of easily-melted,
incompatible element enriched materials in a more refractory, depleted matrix. This result
has implications for the way in which oceanic basalts can be used as probes of mantle
composition, because for a given mantle composition, the incompatible trace element and
isotope compositions of the melts produced will vary with the average degree of melting.
Since most seamount lavas are erupted on the ridge flanks, they are likely to sample preferentially the more enriched, relatively fertile mantle components. Thus to some degree,
the compositions of oceanic lavas may be controlled by lithosphere thickness (Beier et al.,
2011; Ellam, 1992; Haase, 1996; Humphreys and Niu, 2009; Ito and Mahoney, 2005b;
Dissertation P.A. Brandl
63
4. Volcanism on the flanks of the East Pacific Rise
Regelous et al., 2003), without the need to invoke large differences in mantle composition.
Stracke and Bourdon (2009) pointed out that if melts of heterogeneous mantle are pooled
before eruption, then because initial, small degree melts are dominated by the more enriched, fertile component, and because higher degree melts are a mixture of the enriched
and depleted materials, highly depleted melts derived from the depleted component alone
will rarely be erupted at the surface. This is evident from Figure 4.8a; the ‘pooled melt’
compositions extend from high La/Nd, low Yb/Nd ratios characteristic of small degree
melts of the pyroxenite, to lower La/Nd and higher Yb/Nd ratios which nevertheless
do not approach the depleted compositions of high-degree melts of the peridotite component. Melts of depleted mantle lithologies may therefore only be observed in oceanic
basalts in unusual tectonic settings (e.g., Wendt et al., 1999), and the composition of the
upper mantle may be significantly more depleted than commonly assumed from analyses
of oceanic basalts (Stracke and Bourdon, 2009). Information on the composition of the
depleted mantle components may only be obtained from analysis of melt inclusions in
MORB (e.g., Sobolev and Shimizu, 1993), and from abyssal peridotites which represent
the residues of melting beneath spreading ridges (e.g., Johnson et al., 1990; Niu, 2004).
Variable degrees of melting of a heterogeneous mantle results in ‘melt extraction trajectories’ that may lead to misleading conclusions regarding the petrogenesis of the lavas.
For example, unlike simple mixing arrays, these melting trajectories need not necessarily
pass through the endmember mantle compositions (Fig. 4.8a), so that identification of the
origin of ‘recycled materials’ in the source of oceanic basalts based on trace element ratios
may not be straightforward. In addition, models of mantle melting at mid-ocean ridges
that assume that the MORB mantle is homogeneous (Klein and Langmuir, 1987; Salters
and Hart, 1989) are unlikely to be realistic. Haase et al. (2011b; see chapter 5) have shown
that variations in the degrees of melting of a two-component mantle can produce a range
in apparent ‘garnet signatures’ in the erupted lavas without the need to invoke differences
in the fraction of melting taking place in the stability field of garnet peridotite. For example, Bourdon et al. (1996) interpreted the negative correlation between (230 Th/238 U) and
axial ridge depth in global MORB as the result of variations in the proportion of melting
in the stability field of garnet due to regional differences in mantle temperature. However,
if axial depth is controlled on a regional scale by variations in mantle composition (Niu
and O’Hara, 2008), then the uranium-series systematics could instead reflect regional differences in the amount of a fertile, garnet-rich lithology with higher melt productivity,
from which melts with lower (230 Th/238 U) are produced (e.g., Elkins et al., 2008; Russo
et al., 2009; Stracke et al., 1999). Further geochemical and geochronological studies of the
lavas from individual seamounts are needed to quantify these effects in more detail.
64
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Acknowledgements
We are very grateful to R. Batiza for providing sample material and information and
his help to improve this publication. We thank G. Ito and J. J. Mahoney for providing
their MATLAB scripts and help with those, L. Pflug for his help with MATLAB, and
A. Stracke and J. Maclennan for helpful discussions. The manuscript benefitted from the
constructive comments of co-editor K. Mezger and two anonymous reviewers. We thank
the Smithsonian Institution for providing electron microprobe standards, A. Richter for
support during EMP calibration, F. Stöckhert for help with sample preparation, and H.
Beck and DB for inspiration. Furthermore, we are indebted to H. Feldmann, I. Raczek, and
G. Borngässer for their help during isotope analyses, and to J. White and D. Maicher for
providing cruise and submersible data. W. Abouchami was funded by the DFG through
the Leibniz award to K. Mezger.
4.6 Supplementary Information
Parameters of the applied melting model
Table 4.4: Partition coefficients used in Stracke and Bourdon (2009)
Ti
Sr
Y
Zr
Nb
Rb
Ba
La
Ce
Nd
Sm
Eu
Gd
Dy
Er
Yb
Lu
Hf
Ta
Pb
Th
U
Ol
0.015
0.0004
0.0099
0.0033
0.004
0.0003
0.0011
0.0005
0.00025
0.001
0.0013
0.0005
0.0011
0.0027
0.0132
0.0305
0.0432
0.001
0.02395
0.0035
0.00005
0.00089
Opx
0.086
0.0007
0.06175
0.01856
0.00347
0.0002
0.001
0.004
0.00445
0.01309
0.01888
0.009
0.0065
0.065
0.05671
0.09086
0.10857
0.02826
0.00765†
0.0091
0.00135
0.00509
Peridotite
low-Ca Cpx high-Ca Cpx
0.14
0.35
0.091
0.091
0.18186
0.52233
0.04087
0.138
0.01469
0.04567
0.0004
0.0004
0.0008
0.0025
0.015
0.03
0.03121
0.09602
0.058
0.2025
0.08643
0.293
0.115
0.550
0.16
0.35
0.17
0.4
0.17043
0.561
0.22657
0.45
0.23643
0.591
0.07729
0.25525
0.04068
0.1265
0.0071
0.0135
0.005
0.01433
0.00597
0.01315
Gt
0.6
0.0007
2.46533
0.5305
0.02208
0.0002
0.0006
0.0007
0.01963
0.07617
0.26542
0.4
1.2
2.0
2.71875
4.96977
5.98242
0.48342
0.02354
0.005
0.0145
0.03328
Ref.
[1]
[1]
[2][3]
[2][3]
[2][3]
[1]
[2][3]
[1]
[1]
[2][3]
[2][3]
[1]
[1]
[1]
[2][3]
[2][3]
[2][3]
[2][3]
[2][3]
[2][3]
[2][3]
[2][3]
Pyroxenite
Cpx
Gt
0.445
0.345
0.06
0.01167
0.615
3.62
0.13
0.45667
0.008 0.01033
0.003
0.005
0.0055 0.00483
0.0285 0.00683
0.0575 0.00933
0.134
0.2125
0.2515 0.3055
0.217 0.35583
0.4015
0.895
0.535 2.47667
0.68
4.91167
0.765
7.61
0.825
8.97
0.24
0.4
0.0235 0.01167
0.0435 0.03717
0.005 0.00303
0.0064 0.01753
Ref.
[4]*
[4]*
[4]*
[4]*
[4]*
[4][5]
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
[4]*
*partition coefficients are means (Cpx: experiments A343 and MP240; Gt: A343, MP214-240 and MP237254) and corresponds to ‘average pyroxenite’ of Stracke and Bourdon (2009)
Dissertation P.A. Brandl
65
4. Volcanism on the flanks of the East Pacific Rise
†
DT a is average of DN b from [2][3] divided by the DN b /DT a ratio of [6][7]
Note: Spinel is negligible since element concentrations calculated by partition coefficients multiplied by
abundances are almost zero. Ol: olivine, Opx: orthopyroxene, Cpx: clinopyroxene, Gt: garnet, Sp: spinel.
Table 4.5: Mineral mode: Lherzolite (peridotite)
Depth
>75 km
<75 km
Ol
0.53
0.53
Opx
0.04
0.14
Cpx
0.33
0.3
Gt
0.1
0
Sp
0
0.03
Table 4.6: Melting mode: Lherzolite (peridotite)
Depth
>75 km
75-60 km
60-48 km
48-33 km
33-24 km
<24 km
Ol
0.05
0.375
-0.25
-0.45
-0.4
0.2
Opx
-0.49
-0.5
-0.25
0.403
0.6
0.2
Cpx
1.31
1.125
1.5
1.047
0.8
0.5
Gt
0.13
0
0
0
0
0
Sp
0
0
0
0
0
0.1
Table 4.7: Spinel-garnet transition: sign reversed to the rest of the reaction
Cpx
0.6
Gt
-1
Sp
0.4
Table 4.8: Mineral mode: Pyroxenite
Cpx
0.8
Gt
0.18
Qz
0.02
Table 4.9: Melting mode: Pyroxenite
Cpx
0.588
66
Gt
0.229
Qz
0.183
Dissertation P.A. Brandl
4. Volcanism on the flanks of the East Pacific Rise
Table 4.10: Input source composition
Element
Rb
Ba
Th
U
Nb
Ta
La
Ce
Pb
Nd
Sr
Zr
Hf
Sm
Eu
Ti
Gd
Dy
Y
Er
Yb
Lu
Ref.
C0 D-DMM
0.02
0.227
0.004
0.0018
0.0864
0.0056
0.134
0.421
0.014
0.483
6.092
4.269
0.127
0.21
0.086
650
0.324
0.471
3.129
0.329
0.348
0.056
[8]
C0 MORB source (DMM)
0.088
1.2
0.0137
0.0047
0.211
0.0139
0.234
0.772
0.023
0.713
9.8
7.94
0.199
0.27
0.107
798
0.395
0.531
4.07
0.371
0.401
0.0634
[1]
C0 recycled crust
0.59
4.61
0.075
0.021
2
0.13
1.74
6.12
0.099
6.67
78.3
64.7
1.77
2.38
1.05
7838
4.06
5
29.1
3.07
3.2
0.456
[9]
References in Tables
[1] Salters and Stracke (2004); [2] Salters et al. (2002); [3] Salters and Longhi (1999); [4]
Pertermann et al. (2004); [5] Klemme et al. (2002); [6] McDade et al. (2003a); [7] McDade
et al. (2003b); [8] Workman and Hart (2005); [9] Stracke and Bourdon (2009)
Dissertation P.A. Brandl
67
5. Post-spreading volcanism on the fossil Galapagos Rise
5 Insights into mantle composition and
mantle melting beneath mid-ocean
ridges from post-spreading volcanism
on the fossil Galapagos Rise
Karsten M. Haase1,2 , Marcel Regelous1 , Robert A. Duncan3 , Philipp A. Brandl1 , Nicole
Stroncik4 and Ingo Grevemeyer5
1
GeoZentrum Nordbayern, Friedrich-Alexander-Universität Erlangen-Nürnberg, Schlossgarten 5, 91054 Erlangen, Germany
2
previously at Institut für Geowissenschaften der Universität Kiel, Olshausenstr. 40, 24118
Kiel, Germany
3
College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, Oregon 97331-5503, USA
4
GeoForschungsZentrum Potsdam, Telegrafenberg B123, 14473 Potsdam, Germany
5
Leibniz-Institut für Meereswissenschaften an der Universität Kiel, Wischhofstr. 1-3, 24148
Kiel, Germany
Abstract
New major and trace element and Sr, Nd and Pb isotope data, together with 40 Ar/39 Ar
ages for lavas from the extinct Galapagos Rise spreading centre in the eastern Pacific reveal
the evolution in magma compositions erupted during slowdown and after the end of active
spreading at a mid-ocean ridge. Lavas erupted at 9.2 Ma, immediately prior to the end of
spreading are incompatible element depleted mid-ocean ridge tholeiitic basalts, whereas
progressively younger (7.5 to 5.7 Ma), post-spreading lavas are increasingly alkalic, have
higher concentrations of incompatible elements, higher La/Yb, K/Ti, 87 Sr/86 Sr, and lower
143
Nd/144 Nd ratios, and were produced by smaller degrees of mantle melting. The large,
correlated variations in trace element and isotope compositions can only be explained by
melting of heterogenous mantle, in which incompatible trace element enriched lithologies
Dissertation P.A. Brandl
69
5. Post-spreading volcanism on the fossil Galapagos Rise
preferentially contribute to smaller degree mantle melts. The effects of variable degrees
of melting of heterogeneous mantle on lava compositions must be taken into account
when using MORB to infer the conditions of melting beneath active spreading ridges.
For example, the stronger ‘garnet signature’ inferred from Sm/Nd and 143 Nd/144 Nd ratios
for post-spreading lavas from the Galapagos Rise results from a larger contribution from
enriched lithologies with high La/Yb and Sm/Yb, rather than a greater proportion of
melting in the stability field of garnet peridotite. Correlations between ridge depth, and
Sm/Yb and fractionation-corrected Na concentrations in MORB worldwide could result
from variations in mantle fertility and/or variations in the average degree of melting,
rather than large variations in mantle temperature. If more fertile mantle lithologies are
preferentially melted beneath active spreading ridges, then the upper mantle may be
significantly more ‘depleted’ than is generally inferred from the compositions of MORB.
5.1 Introduction
Numerous studies of the lavas erupted at active mid-ocean spreading ridges have shown
that even in the absence of nearby hotspots, the upper mantle is chemically and isotopically heterogeneous. At most spreading ridges, a spectrum of lava compositions is observed, ranging from highly depleted lavas (N-MORB) with low 87 Sr/86 Sr, high
143
Nd/144 Nd, low concentrations of incompatible elements, and lower ratios of more- to
less-incompatible elements (e.g., La/Sm), to rarer, highly enriched mid-ocean ridge basalt
(E-MORB) with high concentrations of incompatible elements, high 87 Sr/86 Sr and low
143
Nd/144 Nd. The enriched isotopic characteristics of E-MORB indicates that these heterogeneities may have ‘ages’ of a few hundred My (e.g., Donnelly et al., 2004). The origin
of these heterogeneities is debated, but they may have an origin in recycled material that
was metasomatised by small degree melts, either in the lowermost oceanic lithosphere
close to spreading ridges, or in the mantle overlying the slab at subduction zones (Niu
et al., 2002; Donnelly et al., 2004; Pilet et al., 2005). The distribution of compositions is
skewed, such that the more enriched E-MORB make up a smaller proportion of the lavas
erupted at most ridges, and the log distribution of concentrations of highly incompatible elements such as Th is approximately normal (Arevalo Jr. and McDonough, 2010).
The enriched material apparently makes up a small volume of the upper mantle, possibly
in the form of ‘veins’ or streaks, but may contain significantly higher concentrations of
the most incompatible elements compared to the enclosing depleted matrix (Batiza and
Vanko, 1984; Sleep, 1984; Zindler et al., 1984; Prinzhofer et al., 1989).
Several lines of evidence suggest that enriched mantle lithologies may have lower solidus
temperature than the more depleted matrix, and so are preferentially tapped at low de-
70
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
grees of melting (Sleep, 1984; Prinzhofer et al., 1989; Ito and Mahoney, 2005a). Lavas
erupted at intra-transform spreading segments in the Garrett Fracture Zone, interpreted
to result from melting of mantle that recently underwent melt extraction beneath the
adjacent spreading ridge, have trace element and isotope compositions that are more
depleted than lavas from neighbouring ridge segments (Wendt et al., 1999). Conversely,
lavas believed to result from small degrees of mantle melting tend to have more enriched
incompatible trace element and isotope compositions (e.g., Haase, 1996; Janney et al.,
2000; Regelous et al., 2003; Hirano et al., 2006; Konter et al., 2009; Castillo et al., 2010).
Clinopyroxenes in residual abyssal peridotites tend to have more radiogenic Nd isotope
compositions than those of lavas from the same section of ridge, consistent with preferential melting-out of eclogite or pyroxenite with lower Sm/Nd and 143 Nd/144 Nd ratios during
decompression melting (Salters and Dick, 2002). Although these observations could be explained by melting of heterogeneous mantle in which enriched lithologies melt to a greater
extent than more depleted lithologies, such ‘non-modal’ melting has not been convincingly
demonstrated. Yet if this process is important during mantle melting beneath spreading
ridges, then melting models that assume that the mantle is homogenous at the scale of
the melting region (e.g., Klein and Langmuir, 1987; Salters and Hart, 1989; McKenzie
and O’Nions, 1991; Spiegelman and Elliott, 1993) are unlikely to be realistic.
Fossil spreading ridges, formed when active spreading centres are abandoned, may continue to erupt magma for several million years after plate separation has ceased (Batiza,
1977; Batiza et al., 1982; Batiza and Vanko, 1985; Davis et al., 2002; Clague et al., 2009;
Haase et al., 2011a). Previous geochemical studies of the youngest lavas erupted at fossil
spreading ridges have shown that many are E-MORB, with highly enriched trace element
and isotope compositions (Batiza and Vanko, 1985; Bohrson and Reid, 1995; Choe et al.,
2007; Choi et al., 2007; Clague et al., 2009; Castillo et al., 2010; Haase et al., 2011a;
Tian et al., 2011). Post-spreading lavas erupted at fossil ridges may therefore preserve the
purest expression of the E-MORB source among all oceanic basalts away from hotspots,
and they apparently result from smaller degrees of mantle melting of the same ‘normal’
mantle that melts beneath actively-spreading ridges to produce MORB. The compositions of post-spreading lavas may therefore give unique insights into the effects of mantle
heterogeneity and the degree of mantle melting on the compositions of melts erupted at
spreading ridges (Batiza et al., 1989; Castillo et al., 2010). On many fossil ridges, the compositions of these post-spreading lavas vary systematically with age at a given location
(e.g., Castillo et al., 2010; Haase et al., 2011a), and may therefore potentially be used to
infer changes in the degree and depth of melting during slowdown and eventual cessation
of spreading.
Here, we present new major and trace element and Sr, Nd, Pb isotope data, and 40 Ar/39 Ar
Dissertation P.A. Brandl
71
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.1: Tectonic map of the eastern Pacific, showing the location of the Galapagos Rise fossil
spreading centre on the Nazca Plate to the east of the active East Pacific Rise. The three fossil ridge
segments immediately south of the South Gallego Fracture Zone (red box) were sampled during RV Sonne
cruise SO-160 (see Fig. 5.2).
ages for lavas erupted during the late spreading and post-spreading stages of the extinct
Galapagos Rise spreading centre.
5.2 Tectonic setting and sample locations
The Galapagos Rise is an extinct (fossil) spreading centre located on the Nazca Plate
in the southeastern Pacific (Menard et al., 1964). Spreading at this ridge, between the
Nazca Plate to the east and the Bauer Microplate to the west, began 18.5 Ma ago, but
was abandoned approximately 12 Ma later when spreading was transferred to the East
Pacific Rise, 900 km to the west (Fig. 5.1). Until spreading ceased, the Galapagos Rise at
10◦ S was a fast spreading ridge with an average spreading rate of 170 mm a−1 . Spreading
is estimated to have slowed dramatically at about 6.5 Ma, and the ridge was finally abandoned at the time of Bauer Microplate capture by the Nazca Plate at 5.8 Ma, based on
bathymetric and magnetic data (Anderson and Sclater, 1972; Herron, 1972; Mammerickx
et al., 1980; Eakins and Lonsdale, 2003).
72
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.2: Bathymetric map of the Galapagos Rise fossil spreading axis between 9◦ 20’S and
11◦ 20’S, showing locations of dredged samples (red
circles). Bathymetric contours in 500 m intervals. A
deep (>4,000 m) axial rift characterises the southern end of the northernmost segment (the Northern
Rift), whereas an axial ridge is present on the two
southern segments (Central and Southern Ridges).
The volcanic ridges show no evidence of tectonic
disturbance, and were therefore constructed after
spreading had ceased, by lavas which filled the axial rift and constructed elongated seamounts which
rise up to 2,500 m above the surrounding seafloor to
within 500 of sealevel. The three samples dated in
this study are from dredges 3DS, 14DS and 19DS.
During RV Sonne cruise SO-160, a detailed bathymetric survey and sample dredging
program was carried out along three ridge segments of the Galapagos Rise, each approximately 50 km in length, between the Dana Fracture Zone to the south and the South
Gallego FZ to the north (Haase and Shipboard Scientific Party, 2002). The southern end
of the northernmost segment is characterised by an axial rift, up to 4,700 m deep, with
a sigmoidal shape (Fig. 5.2). A characteristic nodal deep and an inside corner high occur
at the former ridge-transform intersection at the southern end of this segment (Fig. 5.2).
This type of ridge-transform morphology is typical of slow spreading ridges and is rarely
observed at fast-spreading ridges, confirming that the spreading rate on the Galapagos
Rise slowed dramatically prior to extinction. A similar change in ridge morphology occurred shortly before the abandonment of the Mathematician Ridge (Batiza and Vanko,
1985).
South of the transform offset at approximately 10◦ S, the Galapagos Rise has a very
different morphology. No axial rift is observed, instead an axial ridge is present, which at
its northern end rises about 2,500 m above the surrounding seafloor to within 500 m of
sealevel (Fig. 5.2). The ridge is capped by volcanic cones with heights of several hundred
metres; similar cones occur on the flanks of the ridge and on the surrounding seafloor, and
these are likely the youngest volcanic features of this ridge segment (Haase and Shipboard
Dissertation P.A. Brandl
73
5. Post-spreading volcanism on the fossil Galapagos Rise
Scientific Party, 2002). A second, less pronounced axial high is present immediately to the
south of a small offset of the ridge at 10◦ 55’S (Fig. 5.2). The lack of tectonic disturbance
of these younger ridge-centred volcanic features indicates that extension at the ridge axis
had ceased at the time of their formation. Similar large, post-spreading central volcanoes,
elongated parallel to the former spreading axis and partially covering older structures
have been reported from the Mathematician and Guadelupe fossil ridges in the North Pacific and the extinct Wharton Ridge in the northern Indian Ocean. Some volcanoes built
on fossil ridges may become emergent (for example Guadalupe and Socorro Islands), but
there is no evidence from either volcano morphology or the petrology of dredged samples
that the summit of the large seamount on the Galapagos Rise at 10◦ 24’S was previously
above sealevel.
5.3 Samples and analytical methods
During cruise SO-160, samples of volcanic rock were recovered by dredging from four locations along the deep axial rift in the northern part of the study area, and ten locations
along the shallow axial volcanic ridge to the south (Fig. 5.2). A total of 42 of the freshest
samples (13 from the axial rift, 29 from the axial ridge) were selected for analysis. Most
lavas dredged from the rift are plagioclase, plagioclase ± clinopyroxene, or plagioclase
± olivine phyric, sparsely vesicular basalts, those dredged from the axial ridge and its
flanks are mainly aphyric or sparsely plagioclase phyric, vesicular basalts. Fresh volcanic
glass was present on many samples, and as far as possible, analyses were carried out on
handpicked glasses. Weathered surfaces were removed from whole-rock samples, which
were then coarse-crushed, washed thoroughly in deionised water, and powdered in an
agate mortar.
Major and trace element analyses were carried out at the Institut für Geowissenschaften
at the Universität Kiel. For whole-rock major element analysis, 0.6 g of dried rock powder was mixed with lithium tetraborate and ammonium nitrate, fused to a homogenous
glass bead, and analysed using a Phillips PW1400 XRF spectrometer calibrated against
international rock standards. Magnesium numbers (Mg# = 100 x Mg2+ /(Mg2+ + Fe2+ ))
were calculated assuming FeO = 0.86 x FeOT . Major element compositions of glasses were
determined by electron microprobe (JEOL 8900 Superprobe). For glass analyses, a 12 µm
defocussed beam at 15 nA beam current and 15 kV acceleration was used. The instrument was calibrated against natural glass standards, and precision and accuracy for the
VG-2 standard were better than 1% for all major elements. Trace element concentrations
of both glass and whole-rock samples were determined using an upgraded PlasmaQuad
ICP-MS following the procedure outlined in Garbe-Schönberg (1993). Glass samples were
74
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
washed in ultrapure water in an ultrasonic bath before analysis. Accuracy was checked
using international rock standards (data for BHVO-1 are given for reference in Table A8),
and the external precision for most elements was better than 5%.
Before dissolution for isotope measurements, rock powders were leached for one hour
in hot, ultrapure 6M HCl, then washed thoroughly with ultrapure water. Glass samples were washed but not leached before dissolution. Ion-exchange techniques used for
Sr, Nd and Pb separation are described in Hoernle and Tilton (1991). Sr and Pb isotope measurements were carried out at the GEOMAR Kiel, using a Finnigan MAT 262
thermal ionisation mass spectrometer in static mode. Nd isotope ratios were analysed in
dynamic mode on the same instrument. Fractionation corrections were made assuming
86
Sr/88 Sr = 0.1194 and 146 Nd/144 Nd = 0.7219. Repeat measurements of NBS987 yielded
87
Sr/86 Sr = 0.710218±0.000024 (2SD, n=12). Repeat measurements of the Spex and La
Jolla Nd standards gave 143 Nd/144 Nd values of 0.511710 (n=15) and 0.511827±0.000007
(2SD, n=3) respectively. Data in Table A8 are normalised to values of 0.710250 and
0.511855 for the NBS987 and La Jolla standards. Pb isotope measurements were fractionation corrected using repeat measurements of NBS981 (206 Pb/204 Pb 16.909±0.017,
207
Pb/204 Pb 15.455±0.022, 208 Pb/204 Pb 36.584±0.069), and normalised to the values of
Todt et al. (1996). Pb blanks were negligible (<50 pg).
Three samples were selected for dating using the 40 Ar/39 Ar method. An acid-leached
plagioclase separate from sample 14DS-2, and whole-rock portions of samples 3DS-1
and 19DS-1 were irradiated in the 1MW TRIGA reactor at Oregon State University for
6 hours together with FCT-3 biotite (28.04 Ma) as flux monitor. Details of the analytical
methods used are given in Koppers (2003) and Duncan and Keller (2004). Age plateaus
and isochron ages (Table 5.1) were calculated using software described by Koppers (2002).
5.4 Results
5.4.1 Ar-Ar ages of Galapagos Rise lavas
Sample 3DS-1, from the shoulder of the deep rift at 9◦ 24’S (hereafter the ‘Northern Rift’)
yielded an weighted plateau 40 Ar/39 Ar age of 9.18±0.44 Ma, which represents a maximum
age for the abandonment of spreading, since by analogy with the volcanically active rift
zone of the Mid-Atlantic Ridge the last lavas erupted on this ridge segment were likely
emplaced within the rift floor. On the basis of bathymetric and magnetic data, spreading
at the Galapagos Rise is estimated to have finally ceased at 6.5 Ma (Anderson and Sclater,
1972; Mammerickx et al., 1980; Eakins and Lonsdale, 2003), which may indicate that lavas
within the rift itself span an age range of 2–3 Ma.
Dissertation P.A. Brandl
75
5. Post-spreading volcanism on the fossil Galapagos Rise
Table 5.1: Summary of
Sample
3DS-1
14DS-2
19DS-1
Code
02C9947
02C2931
02C3045
40
Ar/39 Ar data for lavas from the Galapagos Rise Fossil Spreading Centre.†
Sample Type
Whole-rock
Plagioclase
Whole-rock
Weighted plateau
Age (Ma) N % 39 Ar MSWD
9.18±0.44 7
100
1.10
5.66±0.88 6
99.3
0.33
7.50±0.60 7
100
0.50
Total fusion
Age (Ma)
9.43±0.75
6.20±0.90
8.00±0.85
Inverse isochron
Age (Ma) 40 Ar/36 Ar
6.70±3.14 303.5±11.9
4.86±1.79 300.5±9.7
5.63±2.59 299.2±5.2
†
N is number of heating steps, and % 39 Ar indicates percentage of total 39 Ar released used in
plateau age calculation. Total decay constant of 40 K is taken to be 5.530×1010 a−1 . All errors: 2SD.
The two lavas from the volcanic ridge yielded significantly younger 40 Ar/39 Ar ages. Sample
14DS-2, from the western flank of the larger, northern axial high centred at 10◦ 24’S
(‘Central Ridge’) has an age of 5.66±0.88 Ma, whereas sample 19DS-1, which was dredged
from the southern, smaller high along the axial ridge near 11◦ 05’S (‘Southern Ridge’)
yielded an age of 7.50±0.60 Ma. There is no evidence that these younger volcanic features
have been disrupted by faulting, which suggests that active spreading on this part of the
Galapagos Rise ended at between 9.2–7.5 Ma. The youngest of the three ages reported here
was obtained from the largest volcanic construction, and is probably a maximum age for
the youngest flows, since this sample was dredged from the ridge flanks, rather than from
the small cones close to the summit region which are likely to be the youngest volcanic
features. Based on our new ages, magmatism on the Galapagos Rise therefore continued
along part of its length for at least 1.8 Ma after spreading ceased, with post-spreading
lava flows filling the rift and building an axial ridge.
5.4.2 Major and trace element geochemistry
New major and trace element data for Galapagos Rise lavas are listed in Table 1. All
samples recovered from the Northern Rift are tholeiitic basalts, with MgO concentrations
of 6.22 to 8.61 wt. %. The 4 glass samples from dredges 3DS and 6DS lie on well-defined
lines in major element diagrams, whereas whole-rock samples show more scatter, due to
the effects of alteration or variable phenocryst contents (Fig. 5.3). With the exception of
one sample (6DS-1), lavas from the Northern Rift have K2 O/TiO2 and La/Sm ratios of
0.06–0.20 and 0.65–0.97 respectively (Fig. 5.3, 5.4). Their major and trace element compositions therefore lie within the range of ‘normal’ depleted mid-ocean ridge basalts, but
at the depleted end of this range. For example, Nb concentrations of the most depleted
samples are <1 ppm (Fig. 5.4), and the low La/Sm and Nb/Zr ratios in these samples
overlap with the most depleted lavas from near-ridge seamounts on the flanks of the East
Pacific Rise (Fig. 5.4), which have highly variable trace element compositions (Niu and
Batiza, 1997). However, Rb and Ba concentrations are within the range of normal MORB,
and thus Ba/Nb and Ba/Th ratios of the lavas from the Northern Rift are relatively high
(Fig. 5.4).
76
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.3: Major element compositions of Galapagos Rise lavas. Small symbols are for wholerock samples, large symbols for glass analyses. Data for basaltic and andesitic lavas from the northern
East Pacific Rise (black circles; data from Niu et al. (1999) and Regelous et al. (1999)) are shown for
comparison. Compared to lavas from the Northern Rift, younger post-spreading lavas from the Central
and Southern Ridges have lower SiO2 and CaO, and higher Na2 O, K2 O and Al2 O3 for a given MgO.
The post-spreading lavas also have higher Na72 values, and extend to lower Si72 (where Na72 and Si72
are Na2 O and SiO2 concentrations corrected for the effects of fractional crystallisation to Mg#=72 using
the method of Niu et al. (1999). See text for discussion.
Lavas from the Central and Southern Ridges are alkalic basalts with lower MgO concentrations than the lavas from the Northern Rift. Highly evolved lavas, such as trachytes
and rhyolites which have been reported from some post-spreading structures located on
other fossil spreading ridges (e.g., Batiza, 1977; Batiza et al., 1989; Davis et al., 1995) are
not among the samples we have analysed from the Galapagos Rise. However, dredge 15DS
from a cone on the summit recovered a fragment of apparently heavily-altered trachyte
(Haase and Shipboard Scientific Party, 2002). Major element compositions of lavas from
the Northern Rift and from the Central and Southern Ridges overlap, but for a given
MgO, lavas from the latter have lower SiO2 , FeO, CaO, higher Al2 O3 , Na2 O, and significantly higher K2 O (Figure 5.3). Lavas from the Central and Southern Ridges are alkalic
basalts with high concentrations of highly incompatible elements, and higher ratios of
more- to less-incompatible elements, e.g., K2 O/TiO2 , La/Sm, Nb/Zr, compared to lavas
Dissertation P.A. Brandl
77
5. Post-spreading volcanism on the fossil Galapagos Rise
from the Northern Rift (Fig. 5.3, 5.4). The most enriched lavas have Th, Nb concentrations and Nb/Zr, La/Sm ratios that overlap with those of the most enriched East Pacific
Rise (EPR) seamount lavas and extend to higher values (Fig. 5.4). These lavas therefore
include some of the most ‘enriched’ examples of oceanic basalts not associated with longlived intraplate (‘hotspot’) magmatism. A very wide range of magma compositions was
thus erupted at the Galapagos Rise axis within a 2 to 3 Ma period and over a distance
of approximately 150 km; for instance Th and Nb concentrations in Galapagos Rise lavas
vary by over two orders of magnitude (Fig. 5.4) despite a limited range in MgO and Mg#,
and La/Sm and Nb/Zr ratios show a similar range to that found in EPR seamount lavas
(Fig. 5.4), which encompass much of the range observed within MORB worldwide (Niu
and Batiza, 1997).
Figure 5.4: Trace element compositions of Galapagos Rise lavas. Variation of (a) Th and (b)
Ba with Nb concentrations, and (c) Nb/Zr and (d) La/Sm ratios with Nb and La respectively. Lavas
erupted within the Northern Rift during the last stages of active spreading on the Galapagos Rise have
low incompatible trace element concentrations, and low La/Sm and Nb/Zr ratios which overlap with
the range for more depleted lavas from the East Pacific Rise (EPR) and EPR seamounts. In contrast,
post-spreading lavas from the volcanic ridge have higher concentrations of incompatible trace elements,
high La/Sm and Nb/Zr, and are among the most enriched E-MORB found in the ocean basins away from
hotspots. Within the post-spreading lavas there appears to have been a systematic evolution to more
‘enriched’ compositions between 7.5 Ma (Southern Ridge) to 5.7 Ma (Central Ridge). A very wide range
of lava compositions were therefore erupted on the Galapagos Rise within an approximately 2 Ma period.
Data for northern EPR MORB and near-ridge seamount lavas are from Niu and Batiza (1997), Niu et al.
(1999, 2002), Regelous et al. (1999).
78
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
There is evidence that seafloor alteration and weathering may have affected the concentrations of more mobile elements in whole-rock samples. For example, the glass samples
have Nb/U within the range of fresh oceanic basalts (43–47), whereas whole-rock samples
have lower and more variable ratios of between 23 and 33. Whole-rock samples also have
more variable Ce/Pb (11–26) and Ba/Rb (4.3–35; not shown) than those typical of fresh
oceanic basalts (25±5 and 11±3, respectively), and are therefore likely to have gained U,
Rb and Pb. Interestingly, both whole-rock and glass samples have relatively high Ba/Th
and Ba/Nb ratios compared to most MORB, and so the relatively high Ba concentrations
of lavas from the Northern Rift (Fig. 5.3) are therefore apparently a primary feature unrelated to alteration. Nevertheless, in the following discussion we use the less mobile rare
earth and high field strength elements to investigate the petrogenesis of the Galapagos
Rise lavas.
5.4.3 Sr, Nd, and Pb isotope compositions
New Sr, Nd and Pb isotope data for Galapagos Rise lavas are given in Table A8. There
are no systematic differences in isotope composition between glasses and leached wholerock powders, suggesting that any effects of alteration on Sr and Pb isotope compositions
have been removed by the leaching process. All but one of the samples from the Northern Rift have 87 Sr/86 Sr and 143 Nd/144 Nd ratios of 0.70251–0.70264 and 0.51316–0.51321
respectively, and are distinct from lavas from the Central and Southern Ridges (0.70291–
0.70311 and 0.51297–0.51303, see Fig. 5.5). One Northern Rift sample (6DS-1) has an
intermediate Nd composition, and a 87 Sr/86 Sr ratio that lies within the range of samples from the Southern Ridge. This sample also has the highest La/Sm of the Northern
Rift lavas. Lavas from the Central and Southern Ridges and from the Northern Rift
also have different Pb isotope compositions (Fig. 5.5): the latter have less radiogenic Pb
(206 Pb/204 Pb of 17.92–18.09), except for sample 6DS-1 which has a composition within
the range of Ridge lavas (206 Pb/204 Pb of 18.50–18.98). Northern Rift lavas have isotope
compositions within the range of Pacific MORB far from hotspots, although 143 Nd/144 Nd
and 206 Pb/204 Pb ratios lie at the high end and low end of the MORB range, respectively.
The Sr and Nd isotope compositions of lavas from the Central and Southern Ridges overlap with the enriched (high 87 Sr/86 Sr, low 143 Nd/144 Nd) end of the array defined by Pacific
near-ridge seamounts (Fig. 5.5). Lavas from the Galapagos Rise thus display the entire
range in Sr and Nd isotope composition observed on eastern Pacific spreading centres and
near-ridge seamounts away from hotspots.
Our new 40 Ar/39 Ar ages suggest that on the Galapagos Rise there was a systematic
evolution of magmatism from incompatible element depleted, N-MORB-type tholeiitic
basalts towards more alkalic, incompatible element enriched lavas with higher 87 Sr/86 Sr
Dissertation P.A. Brandl
79
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.5: Radiogenic isotope compositions of lavas from the Galapagos Rise, with data
for MORB from the East Pacific Rise and East Pacific Rise (PetDB database) and Pacific near-ridge
seamounts (Niu et al., 2002) shown for comparison. Galapagos Rise lavas cover most of the range in Sr, Nd
and Pb isotope compositions found in Pacific MORB. Excluding sample 6DS-1, post-spreading lavas from
the Central and Southern Ridges have higher 87 Sr/86 Sr, lower 143 Nd/144 Nd ratios and more radiogenic
Pb isotope compositions than older lavas from the Northern Rift which were erupted immendiately before
spreading ceased. Dotted line in (a) and (b) is the Northern Hemisphere Reference Line (Hart, 1984).
and lower 143 Nd/144 Nd after spreading ceased. Although our geochronological data are
limited, a similar temporal evolution in lava compositions has also been reported from
the fossil Phoenix Ridge (Choe et al., 2007; Choi et al., 2007; Haase et al., 2011a), and
fossil spreading centres off Baja California Sur (Tian et al., 2011). On the Galapagos Rise,
the post-spreading lavas with the highest 87 Sr/86 Sr, 206 Pb/204 Pb and lowest 143 Nd/144 Nd
also have the highest incompatible trace element concentrations, and the highest Nb/Zr,
La/Sm and K2 O/TiO2 ratios (Fig. 5.6).
5.4.4 Comparison with lavas from other extinct spreading centres
Geochemical data for lavas from other fossil spreading centres have been reported from
the Mathematician, Guadalupe and other extinct ridges in the NE Pacific, including associated subaerial islands (Batiza and Chase, 1981; Batiza and Vanko, 1985; Clague et al.,
2009; Castillo et al., 2010; Tian et al., 2011), the Antarctic-Phoenix Ridge in the Drake
Passage (Choe et al., 2007; Choi et al., 2007; Haase et al., 2011a), the Wharton Ridge in
80
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.6: Trace element and isotope compositions of lavas from the Galapagos Rise. The
large, correlated variations in incompatible element concentrations, incompatible element ratios, and
isotope ratios cannot be explained by variable degrees of melting of a homogenous mantle source. Data
for lavas from the EPR and Pacific near-ridge seamounts shown for comparison (data sources as for Fig.
5.4).
the northern Indian Ocean (Hébert et al., 1999), as well as the Galapagos Rise (Batiza
et al., 1982). In many cases, the samples analysed are from volcanic features which were
clearly built after spreading ceased.
Based on these previous studies, and our new data for the Galapagos Rise, lavas erupted at
extinct spreading centres have the following geochemical characteristics: they are generally
more alkaline in composition compared to the tholeiites erupted at active spreading centres, and may include relatively evolved lavas such as trachyandesites and trachytes. The
latter difference likely results from the lower magma supply rates beneath fossil spreading ridges (compared to active spreading centres) resulting in longer crustal residence
times and greater degrees of fractionation. Post-spreading lavas from fossil ridges tend to
have higher concentrations of incompatible trace elements, and higher ratios of more- to
less-incompatible elements (La/Sm, Nb/Zr, K2 O/TiO2 ); and generally more radiogenic
Sr and less radiogenic Nd isotope compositions (Fig. 5.7). Post-spreading magmatism on
the Phoenix Ridge, like that on the Galapagos Rise, became increasingly ‘enriched’ with
time (Haase et al., 2011a). To some extent, the major and trace element characteristics of
post-spreading lavas could result from smaller average degrees of melting, resulting from
Dissertation P.A. Brandl
81
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.7: Comparison of trace element and
isotope compositions of post-spreading lavas from
the Galapagos Rise (data from this study) with lavas
from other fossil spreading centres worldwide (data
from Choe et al. (2007), Choi et al. (2007), Castillo
et al. (2010), Haase et al. (2011a)). Post-spreading
lavas include some of the most extreme E-MORB
compositions (high Nb/Zr, La/Sm and 87 Sr/86 Sr)
found far from ‘hotspots’. Data for lavas from the
EPR and Pacific near-ridge seamounts shown for
comparison (data sources as for Fig. 5.4).
less extensive mantle upwelling after spreading ceased. However, the isotopic differences
also indicate a role for source heterogeneity, as discussed below.
5.5 Discussion
5.5.1 Origin of chemical and isotopic variations
Effects of fractional crystallisation and melting processes
There is evidence that lavas from the Northern Rift and from the Central and Southern
Ridges have undergone fractionational crystallisation of different mineral assemblages, and
these differences must be taken into account before attempting to compare differences in
mantle source composition and melting conditions between the two lava suites.
Major element compositions of the Northern Rift lavas lie within the range of EPR MORB,
and the major element variations are consistent with low-pressure fractionation of olivine
+ plagioclase ± clinopyroxene. Both CaO and Al2 O3 decrease with decreasing MgO (the
relatively high Al2 O3 for a given MgO in 3 samples from dredge 3DS is likely due to
plagioclase accumulation in these whole-rock samples). In contrast, within Central and
Southern Ridge lavas, Al2 O3 contents are much higher for a given MgO, and do not vary
systematically with MgO (Fig. 5.3), indicating that plagioclase fractionation was much
less significant. Within these lavas, CaO correlates negatively with Al2 O3 (Fig. 5.3) and
82
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
Sc correlates positively with MgO, consistent with a role for clinopyroxene±olivine fractionation in the lavas from the Central and Southern Ridges. Both minerals are common
phenocryst phases in these lavas. However, the scatter in major element composition indicates that the lavas cannot be related by crystallisation along a single liquid line of
descent from a common parental magma.
The difference in fractionation behaviour between magmas from the Central and Southern
Ridges and from the Northern Rift likely reflects a higher pressure of fractionation for the
former. At higher pressures, the onset of clinopyroxene crystallisation occurs at higher
temperature (Presnall et al., 1979; Tormey et al., 1987), resulting in a decrease in CaO
and CaO/Al2 O3 with decreasing MgO in residual melts, and relatively high Al2 O3 for a
given MgO (e.g., Eason and Sinton, 2006). On a global scale, average pressures of MORB
crystallisation are negatively correlated with ridge spreading rate (Michael and Cornell,
1998), and high-Al2 O3 MORB which are inferred to have undergone extensive clinopyroxene fractionation occur preferentially at ridge segment terminations and dying ridge
segments, where high-pressure crystal fractionation is enhanced due to greater conductive cooling and lower magma supply (Michael and Cornell, 1998; Eason and Sinton, 2006).
Most lavas from the Northern Ridge have lower Na2 O and higher SiO2 for a given MgO
compared to lavas from the Central and Southern Ridges (Fig. 5.3). The subparallel
MgO-SiO2 and MgO-Na2 O arrays defined by syn- and post-spreading lavas indicate that
these differences do not result from the differences in fractionation behaviour discussed
above. The Na2 O and SiO2 concentrations of primitive MORB magmas are both sensitive to the degree of mantle melting (Klein and Langmuir, 1987; Langmuir et al., 1992;
Niu and O’Hara, 2008); melts produced by smaller degrees of mantle melting have higher
Na2 O, lower SiO2 for a given MgO. Central and Southern Ridge lavas have higher Na72
(Na2 O concentrations corrected for the effects of low-pressure fractionation to Mg#=72)
and lower Si72 values than the older Northern Rift lavas and most MORB from normal
spreading centres (Fig. 5.3). These differences can be explained if the younger, postspreading lavas result from smaller degrees of mantle melting. Significant changes in the
average degree of mantle melting are expected during abandonment of a spreading ridge;
the decreasing rate of mantle upwelling and the thickening lithosphere will both tend to
result in a decrease in the average degree of melting with time. The changes in the thermal
regime resulting from the slowdown of spreading on the dying Galapagos Rise therefore
appear to have influenced both primary melt compositions and the subsequent fractional
crystallisation paths of these melts.
Dissertation P.A. Brandl
83
5. Post-spreading volcanism on the fossil Galapagos Rise
The role of mantle heterogeneity
To some extent, the variation observed in incompatible trace element ratios between lavas
from the Northern Rift and from the Central and Southern Ridges could also result from
differences in the average degree and depth of melting. Qualitatively, smaller degrees of
mantle melting at greater average pressure, with a greater proportion of melting occurring
within the stability field of garnet peridotite, could account for the higher incompatible
trace element concentrations, higher Nb/Zr, K2 O/TiO2 , La/Yb and lower Yb/Nd in postspreading lavas from the Central and Southern Ridges, compared to the older lavas from
the Northern Rift. However, the absolute range in concentration of highly incompatible
trace elements (a factor of 100 for Th and Nb, Fig. 5.4) and the large range in incompatible trace element ratios such as Nb/Zr, La/Sm and K2 O/TiO2 , cannot be explained by
any reasonable range in the degree of melting of a homogenous mantle source. Instead, the
correlations between incompatible trace element ratios and Sr, Nd isotope composition
(Fig. 5.5) suggest that much of the variation in the former results from source heterogeneity.
The variably depleted-enriched compositions of MORB erupted on individual spreading
ridge segments is often attributed to mixing of ‘normal’, relatively depleted upper mantle
with more enriched materials (e.g., Schilling et al., 1983; Castillo et al., 2000). However, simple mixing processes are unable to explain the compositional variation within
lavas from the Galapagos Rise. Two-component mixing of melts derived from lithologically distinct enriched and depleted mantle lithologies will result in linear arrays in the
La/Nd–Yb/Nd and La/Yb–Sm/Yb diagrams, whereas the Galapagos-EPR data define
curved arrays (Fig. 5.8). For the same reason, melting of a source composed of two different lithologies, which are mixed in variable proportions (mixing of sources) but which both
melt to the same extent, also cannot explain the observations. Instead, variable degrees
of melting of a two-component mantle in which different lithologies have different trace
element and isotope compositions but also different melting behaviour may best account
for the chemical and isotopic variation within Galapagos Rise lavas, as discussed in more
detail below.
Melting a two-component mantle
Hirschmann and Stolper (1996), Phipps Morgan and Morgan (1999); Phipps Morgan
(2001), Ito and Mahoney (2005a,b), Pearce (2005) and Stracke and Bourdon (2009) have
modelled quantitatively the effects of melting a mantle consisting of two or more chemically and isotopically distinct lithologies having different solidus temperature and melt
productivity. As mantle upwells, these lithologies intersect their solidus temperatures at
different times, and depending on their abundance and melt productivity, the more fer-
84
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
Figure 5.8: Results from modelling two-component mixing of melts. Variation of (a) Yb/Nd
and (b) 143 Nd/144 Nd with La/Nd, and 143 Nd/144 Nd with (c) Yb/Nd and (d) δSm/Nd for Galapagos Rise
lavas. δSm/Nd is defined as δSm/Nd = ((Sm/Nd)2Ga - (Sm/Nd)m )/(Sm/Nd)2Ga , where (Sm/Nd)2Ga is
the calculated source Sm/Nd based on the measured present-day 143 Nd/144 Nd ratio of the sample and
assuming a 2 Ga mantle model age, and (Sm/Nd)m is the measured Sm/Nd ratio (Salters, 1996). Twocomponent mixing of melts or sources result in linear arrays on these diagrams; the curved arrays defined
by the Galapagos Rise lavas in (a) to (c) therefore cannot be explained by simple mixing of melts or
by melting of two-component mantle in which enriched and depleted lithologies are mixed in variable
proportions but contribute equally to melting. The negative correlation of 143 Nd/144 Nd with δSm/Nd is
unexpected if δSm/Nd values are controlled only by the proportion of melting within the stability field
of garnet. Instead, the trace element and isotope variations in Galapagos Rise lavas can be reproduced
by variable degrees of melting of a mantle source consisting of different lithologies which melt at different
rates. Curves show range in pooled melt compositions produced by variable degrees of fractional melting
(residual porosity 0.1%) of a two-component mantle source, calculated using the method of Stracke and
Bourdon (2009). In this model, the source mantle contains ‘veins’ of a volumetrically minor lithology
(‘pyroxenite’) which has higher incompatible trace element concentrations and higher La/Yb, Sm/Yb
and lower Yb/Nd and 143 Nd/144 Nd than the enclosing peridotite matrix. The ‘pyroxenite’ has a lower
solidus temperature and therefore contributes more to melting at low melt fractions, compared to the
more refractory matrix, which begins melting at a slightly lower pressure. In a–d, the resulting melt
evolution paths are curved because with increasing degree of melting, the contribution to the pooled melt
from the more fertile component with high incompatible trace element concentrations and high La/Nd,
low Yb/Nd and 143 Nd/144 Nd progressively decreases. Melting functions, source mineralogy and partition
coefficients are taken from Stracke and Bourdon (2009); numbers on the melting curves in a–d indicate
the depth to the top of the melting column. Given the number of variables in these melting models, we
have not attempted to adjust these parameters so as to perfectly reproduce the Galapagos Rise dataset;
rather the purpose of the modelling is to show that, in contrast to binary mixing of endmember melt
compositions, melting of heterogeneous mantle can produce curved arrays on these diagrams. Data for
lavas from the EPR and Pacific near-ridge seamounts shown for comparison (data sources as for Fig. 5.4).
tile components may become exhausted before reaching the top of the melting column.
As they melt, more fertile lithologies extract heat from the surrounding more refractory
Dissertation P.A. Brandl
85
5. Post-spreading volcanism on the fossil Galapagos Rise
lithologies, inhibiting melting of the latter; once more refractory lithologies do begin to
melt, the remaining fertile materials may stop melting (Phipps Morgan, 2001). A consequence of this behaviour is that the resulting melting paths in figures such as 5.8a–d
may be curved, kinked or discontinuous, as different lithologies begin to contribute to
the instantaneous melt composition or become exhausted with increasing melt fraction.
Another feature of such ‘melt extraction trajectories’ (Phipps Morgan and Morgan, 1999)
is that because less than the total number of lithologies may undergo melting at any
time, the range in composition of the melts produced with progressive melting is greater
than would be the case if all lithologies contributed equally to the melt, although the
pooled melts extracted from the melting column will generally have compositions that
are intermediate between the endmember lithologies (Phipps Morgan and Morgan, 1999;
Pearce, 2005; Stracke and Bourdon, 2009). In addition, Ito and Mahoney (2005a,b) have
shown that differences in the mantle flow field within the melting region have a significant
influence on the relative amount of melt that is extracted from enriched and depleted
lithologies.
As discussed above, the major and trace element compositions of Galapagos Rise lavas
together with the new 40 Ar/39 Ar ages indicate that the youngest lavas result from smaller
degrees of mantle melting. During abandonment of a spreading ridge, the average degree
of mantle melting is in fact expected to progressively decrease with time as the mantle
upwelling rate slows and the overlying lithosphere thickens by conductive cooling (Choe
et al., 2007; Choi et al., 2007). Lavas produced by lower degrees of melting of heterogeneous
mantle will contain a larger contribution from more fertile lithologies which are expected to
have higher incompatible element concentrations, higher 87 Sr/86 Sr and lower 143 Nd/144 Nd
(e.g., Pearce, 2005; Stracke and Bourdon, 2009). We have therefore used a forward modelling approach and the melting equations of Stracke and Bourdon (2009) in order to
examine whether variable degrees of melting of heterogeneous mantle can reproduce to
first order the trace element and isotope variations within the Galapagos Rise lavas. We
assume the simplest case of a two-component mantle, consisting of a volumetrically-minor,
incompatible element enriched component with high concentration of incompatible trace
elements, high La/Yb and Nd/Yb, and higher 87 Sr/86 Sr, lower 143 Nd/144 Nd which has a
lower solidus temperature and melt productivity than the surrounding more refractory
lithology. We examine the range of melt compositions produced during variable degrees
of melting of this source material using the melting model of Stracke and Bourdon (2009).
The results are shown in Figure 5.8. Although the combination of model parameters
we have used to calculate the melt extraction paths in Figure 5.8 are non-unique, the
results of the modelling do show that variable degrees of melting of a two-component
mantle can explain many aspects of the geochemistry of Galapagos Rise lavas. In partic-
86
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
ular, such models can reproduce the large variations in the ratios of highly incompatible
elements, and the correlated variations in incompatible trace element and isotope ratios,
which cannot be explained by melting of a homogenous mantle source. At progressively
smaller degrees of melting, fertile enriched materials increasingly dominate melt compositions, and the resulting melt evolution trajectories can reproduce the curved data arrays
defined by Galapagos Rise lavas in Figure 5.8. A similar temporal evolution of lava chemistry is observed in post-spreading lavas from the extinct Phoenix Ridge in the Drake
Passage (Choe et al., 2007; Choi et al., 2007; Haase et al., 2011a), and the fossil spreading
centres off Baja California Sur (Tian et al., 2011). The chemical and isotopic evolution
of post-spreading lavas from fossil ridges may therefore represent some of the strongest
evidence that the ‘normal’ mantle beneath spreading ridges is highly heterogeneous on a
length scale that is small relative to the size of the melting region. At actively spreading
ridges, the average degree of melting is greater and melts are more efficiently mixed on
their way to the surface, with the result that the composition of lavas erupted at active
spreading ridges will provide a less accurate record of the degree of mantle heterogeneity.
Nevertheless, the effects of source heterogeneity must be taken into account when using
MORB to infer the composition of the mantle, as discussed in section 5.5.3.
The highly variable Galapagos Rise lavas were erupted within a period of about
3.5 Ma, over a distance of 150 km, confirming the view that the upper mantle away
from ‘hotspots’ is highly heterogeneous. Thus at least on fossil ridges, large seamounts
composed of E-MORB can apparently be produced by small degrees of melting of ‘normal’
upper mantle, in the absence of any nearby hotspot or plume.
5.5.2 Mantle upwelling and melting beneath spreading ridges
Lavas from fossil ridges provide an opportunity to examine the processes of mantle melting
during slowdown of spreading and in the absence of plate separation, and can potentially
give insights into the nature of mantle upwelling beneath actively-spreading ridges. The
degree to which mantle upwelling beneath mid-ocean ridges is ‘active’ rather than merely
a passive response to plate separation has been extensively debated. Passive upwelling of
mantle is an expected consequence of plate separation (Spiegelman and McKenzie, 1987;
Phipps Morgan et al., 1987), and the wide melt-containing zone beneath the EPR identified during the MELT experiment is consistent with upwelling driven by plate separation
(The MELT Seismic Team, 1998). However, some degree of active upwelling is predicted
to result from ‘melting-induced buoyancy’, due to thermal expansion and the presence of
a melt phase and less dense residual peridotite (Sotin and Parmentier, 1989; Parmentier
and Phipps Morgan, 1990). It has also been proposed that variable, active mantle upwelling beneath mid-ocean ridges augments the passive upwelling and is responsible for
Dissertation P.A. Brandl
87
5. Post-spreading volcanism on the fossil Galapagos Rise
ridge segmentation, with more active deep upwelling and greater melt production beneath
ridge segment centres (Macdonald et al., 1988; Buck and Su, 1989; Scott and Stevenson,
1989; Lin et al., 1990).
On most fossil ridges, the transition from normal spreading to no spreading apparently
took place within about 1 Ma (Batiza, 1989), whereas post-spreading magmatism occurs
over timescales of 2-9 Ma after spreading ceased (Batiza and Vanko, 1985; Batiza, 1989;
Bohrson and Reid, 1995; Choe et al., 2007; Clague et al., 2009; Haase et al., 2011a; this
study). Once formed, melt is efficiently extracted from the mantle within 103–105 years
(e.g., Stracke et al., 2006), and so the age range of post-spreading lavas requires a process
that can actively generate melt over a period of up to 9 Ma after spreading ends. Castillo
et al. (2010) proposed that beneath the thickening oceanic lithosphere at a fossil spreading centre, melting in the absence of plate separation may result from (a) residual mantle
upwelling due to the combined buoyancy effects of thermal expansion, melt depletion and
the presence of small melt fractions, or (b) melting of fertile lithologies as the thinner
lithosphere at the fossil ridge drifts over previously-undepleted mantle. At a fossil ridge,
conductive cooling will erase significant differences in lithosphere thickness over a period
of several tens of Ma after spreading ceases. If post-spreading magmatism results from
upwelling due to variations in lithosphere thickness, magmatism might be expected to
continue over similar timescales. The observed age range of post-spreading magmatism
on fossil ridges is therefore consistent with melting resulting from mantle upwelling due
to variations in lithosphere thickness. However, if lithosphere thickness controls the location of post-spreading magmatism, melting would be expected to occur preferentially
at transform offsets, where differences in lithosphere thickness are most pronounced, and
where upwelling resulting from movement of the lithosphere over the upper mantle would
be concentrated. Batiza (1989) stated “. . . at the Mathematician and Guadelupe failed
rifts, abundant post-abandonment alkalic volcanism is found at failed rift-transform intersections . . . ”, but this does not appear to be the case along the Galapagos Rise (Eakins
and Lonsdale, 2003), nor along sections of other fossil ridges which have been mapped
in detail. It is possible that Batiza and Vanko (1985) may have mislocated the ridge
axis along several of these fossil spreading centres due to the limited bathymetric data
available at that time (Tian et al., 2011). More recent studies have shown that there is
apparently a tendency for post-spreading magmatism to construct axial seamounts away
from segment ends (Choe et al., 2007; Haase et al., 2011a; Tian et al., 2011). On the Galapagos Rise, bathymetric, magnetic and altimetric data show that elongate seamounts are
present along much of the fossil ridge axis, with smaller, isolated seamounts nearer to
segment ends and close to transforms (Eakins and Lonsdale, 2003). There is no evidence
for significant post-spreading magmatism on the ridge segments immediately north or
south of the large-offset South Gallego Fracture Zone, where young, thin lithosphere is
88
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
juxtaposed against much older, thicker lithosphere.
On the Galapagos Rise, the discontinuous distribution of post-spreading magmatism, and
its apparent concentration at the former ridge axis and closer to segment centres is more
consistent with an origin from active mantle upwelling, either resulting from a component
of ‘active’ 3D mantle flow or driven by the buoyancy of melt-depleted residual mantle beneath the ridge. In the latter case, the volume and duration of post-spreading magmatism
might be related to the degree of prior melt depletion. Haase et al. (2011a) observed a
possible correlation between the volume of post-spreading volcanoes on fossil ridges and
the former spreading rate of the ridge. If the degree of mantle melting beneath spreading
ridges is related to spreading rate (Niu and Batiza, 1997), such a correlation may indicate
that melting-induced buoyancy is important. On the other hand, the duration of postspreading magmatism at fossil ridges (up to 8 Ma on Davidson Seamount; Clague et al.,
2009) may be longer than can be accounted for by density contrasts resulting from prior
melting, and indicate that a component of active, 3D upwelling is responsible for continued magmatism on fossil ridges. The distribution and age range of post-spreading lavas
on fossil ridges therefore suggest that active upwelling contributes to the passive mantle
upwelling beneath actively-spreading ridges, especially close to segment centres. The lack
of significant post-spreading volcanism on the Northern Rift may result from less intense
upwelling close to the major South Gallego Fracture Zone, or from less fertile mantle
beneath this ridge segment, as also proposed to explain the discontinuous distribution of
post-spreading volcanism along fossil spreading centres elsewhere (Castillo et al., 2010).
5.5.3 Implications for chemical and isotopic variation in global
MORB
We have shown that the range in incompatible trace element ratios such as La/Yb, Nb/Zr
and K2 O/TiO2 in Galapagos Rise lavas is dominantly the result of variations in the degree
of melting of a heterogeneous mantle. In this location, the effects of source heterogeneity
are relatively large, due to the decreasing degrees of melting resulting from slowdown
and cessation of spreading. Similarly heterogeneous mantle likely underlies much of the
global spreading system, as indicated by the similarity of post-spreading lavas on fossil
ridges worldwide (Fig. 5.7), and highly-variable lavas erupted on seamounts on the flanks
of spreading ridges (Batiza and Vanko, 1984; Zindler et al., 1984; Niu and Batiza, 1997;
Niu et al., 2002). The larger degrees of mantle melting and more complete magma mixing
effects beneath active spreading ridges result in less chemical and isotopic variation within
MORB erupted at active ridges. However, variations in the degree of melting resulting
from differences in the degree of upwelling or mantle temperature, or variations in the
relative volumes and compositions of enriched and depleted lithologies could exert an in-
Dissertation P.A. Brandl
89
5. Post-spreading volcanism on the fossil Galapagos Rise
fluence on MORB chemistry both within individual ridge segments and between different
sections of ridge. As a result, it may not be straightforward to infer the average depth and
degree of melting beneath mid-ocean ridges based on the trace element variation within
MORB.
Nevertheless, numerous studies have attempted to estimate the approximate depth and
degree of mantle melting beneath spreading ridges from the trace element compositions
of MORB. For example, Salters (1996) found that MORB erupted on deeper sections of
the MOR system tend to have a stronger ‘garnet signal’ as inferred from the difference
between Lu/Hf and Sm/Nd ratios of MORB and time-integrated source ratios inferred
from 176 Hf/177 Hf and 143 Nd/144 Nd. This observation apparently conflicts with the major element systematics of MORB, which have been interpreted to indicate that beneath
shallow ridge segments the mantle is hotter and begins melting deeper, so that a greater
proportion of the melt is generated within the stability field of garnet (Klein and Langmuir, 1987; Langmuir et al., 1992). Salters (1996) therefore proposed that beneath deeper
ridges the melting region is broader at its base, such that a greater proportion of melting
occurs within the stability field of garnet. Beneath shallow ridges the melting region is
inferred to extend to greater depths, but because it is columnar in shape a smaller proportion of the total melt is generated within the stability field of garnet (Salters, 1996). In
contrast, Shen and Forsyth (1995) argued that the variation in the apparent garnet signature is predominantly due to variations in the final depth (uppermost limit) of melting,
which lies at greater depth beneath deeper ridge segments. A deeper final depth of melting
could result from higher conductive cooling to the surface, or a lesser degree of mantle
upwelling. Both parameters are expected to be affected by spreading rate, but there is no
simple relationship between spreading rate and ridge depth. Another explanation for the
Sm/Yb-depth relationship (Shen and Forsyth, 1995) is that deeper ridges are underlain
by more fertile, garnet-rich mantle, which upwells more slowly due to its higher density,
thus causing melting to cease at a greater depth (Niu and O’Hara, 2008). In this model,
the melts produced at deep ridges have higher Sm/Yb because they are less diluted by
melts of more refractory peridotite with low Sm/Yb (Niu and O’Hara, 2008).
Our new data for young post-spreading lavas from the Galapagos Rise show that variations in the extent of melting of ‘normal’ heterogeneous mantle have a very significant
effect on the La/Yb, Sm/Nd ratios (and hence inferred garnet effect) of lavas erupted
in this location. Salters (1996) and Shen and Forsyth (1995) attempted to correct for
the effects of mantle heterogeneity on the REE compositions of MORB, but our results
suggest that this may not always be successful. For example, Salters (1996) calculated
δSm/Nd values for MORB, a measure of the difference between the measured Sm/Nd
ratio in a lava and the time-integrated source Sm/Nd ratio inferred from Nd isotope
90
Dissertation P.A. Brandl
5. Post-spreading volcanism on the fossil Galapagos Rise
compositions, and these δSm/Nd values were assumed to be related to the magnitude
of the garnet effect during melting. However, δSm/Nd values of Galapagos Rise lavas
are correlated negatively with 143 Nd/144 Nd ratios (Fig. 5.8), a relationship which is not
expected if variations in δSm/Nd result only from the proportion of melting within the
stability field of garnet. The compositional range of the Galapagos Rise lavas is similar
to that of lavas from near-ridge seamounts on the flanks of the EPR, suggesting that a
similarly heterogeneous mantle is widespread beneath the Pacific far from hotspots. The
correlations of δSm/Nd and Sm/Yb with Na8 , Na72 and axial depth for spreading ridges
worldwide (Salters, 1996; Shen and Forsyth, 1995) may therefore result from variations in
either the average degree of melting of heterogeneous mantle, or the relative proportions
of enriched to depleted lithologies, rather than the depth of melting relative to the spinelgarnet peridotite transition. Beneath deep ridges the average degree of melting is smaller,
and/or the mantle is more ‘enriched’, so that a fertile component with high Sm/Yb contributes more to the total melt (Niu and O’Hara, 2008). Lower degrees of melting at
greater average depth beneath deep ridges could result from a lower degree of mantle
upwelling (Shen and Forsyth, 1995; Niu and O’Hara, 2008), or possibly from greater loss
of heat to the surface due to hydrothermal circulation, which is predicted to penetrate to
greater depths at deep ridges due to the higher hydrostatic pressure (Kasting et al., 2006).
If more ‘enriched’ mantle lithologies preferentially contribute to melting beneath active
spreading ridges, then the isotopic composition of MORB will not faithfully record that
of the upper mantle, as is commonly assumed. Instead, MORB compositions may be biased towards more radiogenic Sr, Pb and Os, and less radiogenic Nd and Hf compositions
(e.g., Phipps Morgan and Morgan, 1999; Stracke and Bourdon, 2009), and a complementary ‘hidden’ depleted component will be retained in the melting residues. 143 Nd/144 Nd
ratios of clinopyroxenes from abyssal peridotites extend to higher values than associated
MORB (Snow et al., 1994; Salters and Dick, 2002; Cipriani et al., 2004; Warren et al.,
2009) and Os isotope compositions of abyssal peridotites are less radiogenic than those of
MORB (Harvey et al., 2006; Liu et al., 2008); accurate Hf and Pb isotope data for abyssal
peridotites are needed to confirm this effect.
5.6 Summary and conclusion
We used major and trace element and Sr, Nd and Pb isotope data, together with 40 Ar/39 Ar
ages for lavas from the Galapagos Rise in the eastern Pacific, to investigate the evolution
in magma compositions erupted during slowdown and after the end of active spreading
on this fossil mid-ocean ridge. Magmatism on the Galapagos Rise continued for at least
2 Ma after active spreading ceased, and younger post-spreading lavas are more alkalic,
Dissertation P.A. Brandl
91
5. Post-spreading volcanism on the fossil Galapagos Rise
have higher concentrations of incompatible elements, higher La/Yb, K/Ti, 87 Sr/86 Sr, and
lower 143 Nd/144 Nd ratios than lavas inferred to have erupted immediately before spreading ended. The very large range in trace element and isotope compositions cannot be
explained by melting of a homogenous mantle source, or by two-component mixing of
enriched and depleted endmember melts, or by melting of variably-enriched mantle in
which both enriched and depleted lithologies contribute equally to melting. Instead, the
trace element and isotope variations can be produced by variable degrees of melting of
a two-component mantle in which incompatible trace element enriched lithologies with
lower melting temperature preferentially contribute to the melt at low degrees of melting.
Post-spreading lavas from the Galapagos Rise therefore contain a greater contribution
from enriched mantle lithologies with lower Sm/Nd and 143 Nd/144 Nd, which yield melts
with higher δSm/Nd, and do not necessarily require a greater proportion of melting in
the stability field of garnet peridotite. Our results, combined with those from other fossil
spreading centres and seamounts on the flanks of spreading ridges, provide clear evidence
for a significant influence of variable degrees of melting of heterogeneous mantle on the
chemical variation in lavas erupted at spreading ridges away from hotspots. This effect
must be taken into account when using the compositions of MORB to infer the conditions of melting beneath active spreading ridges. We suggest that the correlations between ridge depth and δSm/Nd, Sm/Yb and fractionation-corrected Na concentrations in
lavas for actively-spreading ridges worldwide may result from variations in mantle fertility
and/or variations in the average degree of melting, rather than large variations in mantle
temperature. If more enriched mantle lithologies with low 143 Nd/144 Nd, high 87 Sr/86 Sr
are preferentially melted during mantle upwelling beneath active spreading ridges, then
the upper mantle may have significantly higher 143 Nd/144 Nd, lower 87 Sr/86 Sr and a less
radiogenic Pb isotope composition than is commonly inferred from analyses of MORB.
Acknowledgements
We are grateful to Captain Andresen and his crew for their help during Sonne cruise
SO-160, and D. Garbe-Schönberg, C. Voigt, F. Hauff and B. Mader for helping with the
analytical work. G. Ito and A. Stracke very kindly provided us with copies of their mantle
melting models and advised us on their use. We thank C. Beier and A. Stracke for useful
discussions, and the two journal reviewers for their helpful comments. This study was
funded by the Bundesministerium für Bildung und Forschung through grant 03G0160A.
92
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
6 Compositional variation of lavas from
a young volcanic field on the
Southern Mid-Atlantic Ridge, 8◦48’S
Karsten M. Haase1,2 , Philipp A. Brandl1 , Bernd Melchert3† , Folkmar Hauff3 , Dieter
Garbe-Schönberg2 , Holger Paulick4‡ , Thomas Kokfelt3∗ and Colin W. Devey3
1
GeoZentrum Nordbayern, Friedrich-Alexander-Universität Erlangen-Nürnberg, Schlossgarten 5, 91054 Erlangen, Germany
2
Institut für Geowissenschaften der Christian-Albrechts-Universität, Kiel, Olshausenstr.
40, 24118 Kiel, Germany
3
GEOMAR, Helmholtz-Zentrum für Ozeanforschung Kiel, Dienstgebäude Ostufer, Wischhofstr. 1-3, 24148 Kiel, Germany
4
Mineralogisches und Petrologisches Institut, Universität Bonn, Poppelsdorfer Schloss,
53115 Bonn, Germany
†
present address: EEIG Heat Mining, Route de Soultz, F-67250 Kutzenhausen, France
present address: Boliden Mineral AB, 93681 Boliden, Sweden
∗
present address: Geological Survey of Denmark and Greenland, GEUS, Øster Voldgade
10, 1350 Copenhagen K, Denmark
‡
Abstract
Volcanic eruptions along the mid-oceanic ridge system are the most abundant signs of
volcanic activity on Earth but little is known about the timescales and nature of these
processes. The main parameter determining eruption frequency as well as magma composition appears to be the spreading rate of the mid-oceanic ridge. However, few observations
on the scale of single lava flows exist from the slow-spreading Mid-Atlantic Ridge so far.
Here, we present geological observations and geochemical data for the youngest volcanic
features on the slow-spreading (33 mm a−1 ) southern Mid-Atlantic Ridge at 8◦ 48’S. Sidescan sonar mapping revealed a young volcanic field with high reflectivity that was probably
Dissertation P.A. Brandl
93
6. Compositional variation of lavas from a young volcanic field
erupted from two volcanic fissures each of about 3 km length. Small-scale sampling of the
young lava field at 8◦ 48’S by Remotely Operated Vehicle (ROV) and wax corer shows
three different lava units along an about 11 km long portion of the ridge. Based on the
incompatible element compositions of volcanic glasses we can distinguish two lava units
forming the northern and the southern part of the lava field covering an area of 5 and
9 km2 , respectively. Basalts surrounding the lava field and from an apparently old pillow
mound within the young flows are more depleted in incompatible elements than glasses
from the young volcanic field. Radium disequilibria suggest that most lavas from this volcanic field have ages of 3–5 ka whereas the older lavas surrounding the lava field are older
than 8 ka. Faults and a thin sediment cover on many lavas support the ages and indicate
that this part of the Mid-Atlantic Ridge is in a tectonic rather than in a magmatic stage.
Lavas from the northern and southern ends of the southern lava unit have lower MgO but
higher Cl/K than those from the centre of the unit indicating an increased cooling and
assimilation of hydrothermally altered material during ascent, most likely at the tips of
the feeder dike. The compositional heterogeneity on a scale of 3 km suggests small magma
batches that rise vertically from the mantle to the surface without significant lateral flow
and mixing. Thus, the observations on the 8◦ 48’S lava field are in agreement with the
model of low frequency eruptions from single ascending magma batches that has been
developed for slow-spreading ridges.
6.1 Introduction
Volcanic eruptions along the mid-oceanic ridge system represent the volumetrically most
important volcanic process on Earth. However, little is known about eruptive processes on
the seafloor, the magma transport in the crust and during eruption, and the composition
of lavas from single eruptive events. Spreading rate appears to be an important factor in
governing both magma chemistry and the relative importance of magmatic and tectonic
processes for accommodating the plate separation. For example, at slow-spreading axes
the lavas appear to be more primitive, melt lenses occur, if at all, deeper and the chemical
variability appears to be larger, indicating less efficient mixing processes (Rubin et al.,
2009). Closely related mid-ocean ridge lava flows, in some cases from one eruptive event,
have mainly been studied on the East Pacific Rise and on Iceland and have been shown
to consist of lavas with variable compositions (Hall and Sinton, 1996; Perfit and Chadwick, 1998; Sigmarsson et al., 1991; Sinton et al., 2002). Few studies exist on small-scale
compositional variations of lavas on slow-spreading ridges (Stakes et al., 1984; Sinton
et al., 2002) and little is known about the temporal evolution of volcanism (Rubin and
Macdougall, 1990; Sturm et al., 2000). The heterogeneity of melt inclusions within single
olivine and plagioclase crystals as well as the extreme compositional variation and zoning
94
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
Figure 6.1: Bathymetric map of the Southern Mid-Atlantic Ridge (axis shown red) between
Ascension and Bode Verde Fracture Zone and the 2nd order segments A1 to A4. The study area (yellow)
is located on segment A2.
of minerals suggests mixing of highly variable magmas in the magmatic systems (Dungan
and Rhodes, 1978; Shimizu, 1998). An understanding of the composition of single eruptive units could provide insights into the magma transport from the mantle or crust to
the surface. Seismic tomographic studies of slow-spreading axes have revealed complex
magma feeding systems beneath spreading segments where different magma reservoirs
appear to be connected (Magde et al., 2000). One model of the plumbing system suggests
that magma ascends from the mantle preferentially in the centre of a segment and flows
then laterally towards the segment ends, either in the lower crust or through shallower
dikes (Abelson et al., 2001; Magde et al., 2000). Other authors suggest that melt is produced and ascends beneath the whole segment, at least at fast-spreading segments and
slow-spreading segments with high magma production (Tucholke et al., 1997).
Here, we present observations and geochemical data on one eruptive unit of the slowspreading southern Mid-Atlantic Ridge and show that significant compositional variations exist within large lava units. We find two approximately 3 km long volcanic fissures
Dissertation P.A. Brandl
95
6. Compositional variation of lavas from a young volcanic field
apparently fed by dikes erupting two lava units with significantly different incompatible
element compositions.
6.2 Geological setting
This paper presents volcanological and geochemical results on lavas from a relatively
young lava field on the southern Mid-Atlantic Ridge (MAR) at about 8◦ 48’S (Fig. 6.1)
where the full spreading rate is about 33 mm a−1 (DeMets et al., 1994). Previous work
had divided the axis in this region into four second-order segments A1 to A4 (Bruguier
et al., 2003; Fig. 6.1). The water depth decreases from about 3,500 m on segment A1 to
about 2,500 m on segment A2 to only 1,500 m on segment A3 before returning to 3,500m
on segment A4 (Fig. 6.1). Seismic and gravimetric data indicate that crustal thickness
also increases from 5 km on segment A1 to about 10 km in the centre of segment A2
(Bruguier et al., 2003; Minshull et al., 1998). The shallow segments A2 and A3 also do
not show the deep axial rift typical of slow-spreading axes but rather have narrow volcanic
ridges resembling fast-spreading axes. These segments appear to be ‘magmatically robust’
(Scheirer and Macdonald, 1993) and to be significantly more magmatically active than
the deeper segments to north and south. Geochemical and geophysical data suggest that a
melting anomaly underlies segments A2 and A3 because these two segments are unusually
shallow, have a thickened crust and erupt lavas with incompatible element-enriched and
radiogenic Sr and Pb isotopic composition (Hanan et al., 1986; Hoernle et al., 2011;
Minshull et al., 1998).
6.3 Sampling and analytical methods
6.3.1 Sampling and observations
During Meteor cruise M62/5 in November 2004 the southern MAR was mapped using
the Towed Ocean Bottom Instrument (TOBI) side-scan system (Devey et al., 2005). A
large field of young lavas was observed at 8◦ 48’S close to the centre of segment A2 (Fig.
6.2). The high and homogeneous reflectivity of the field suggests approximately similar
ages of eruption for the lavas of this field. The A2 volcanic field was sampled in detail
during the Meteor 64/1 cruise in April 2005 using the MARUM QUEST4000 ROV and a
wax corer. The ROV was navigated using a self-calibrating acoustic IXSEA GAPS USBL
positioning system allowing positioning accurate to within ca. 1% of water depth (± 23 m
at the depths considered here). Lava samples were taken with the hydraulic arms of the
ROV. Previously, three dredge sites had recovered basaltic samples from this field. Some
30 new samples containing fresh volcanic glass were recovered from 20 stations including
two ROV transects across and along the field (Fig. 6.2a).
96
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
Figure 6.2: Details maps of
the 8◦ 48’S volcanic field using combined bathymetric and
sidescan sonar data. (a) Bathymetric map of the lava field and
segment A2 showing the sample
sites. (b) TOBI side-scan sonar
map. (c) Map with the interpretation of tectonic and volcanic structures based on the bathymetric and
sidescan observations.
Dissertation P.A. Brandl
97
6. Compositional variation of lavas from a young volcanic field
6.3.2 Geochemical analyses of glasses
The glass was separated from the samples by hand-picking and cleaned using deionised
water. Glass chips were polished and analysed for their major element composition on a
JEOL JXA–8900 Superprobe electron microprobe at the Institut für Geowissenschaften of
Christian-Albrechts-Universität (Kiel) using standard wavelength-dispersive techniques.
The instrument was operated at an accelerating voltage of 15 kV and beam current of
20 nA. The beam diameter during calibration and sample measurement was 12 µm. Counting times on peaks and background varied depending on the element analysed, and were
20 seconds for all major elements except Na2 O which was analysed with peak counting
times of 10 seconds. Background counting times were always half of peak counting times.
Individual glass chips were analysed at several places and the average was calculated.
Representative glass samples were analysed for trace elements by ICPMS following the
general procedure described previously (Garbe-Schönberg, 1993). Most of the trace element data were published previously (Hoernle et al., 2011). Radiogenic isotope data are
also from Hoernle et al. (2011) where analytical methods can be found. Compiled data
can be found in Table A9 of the Appendix.
6.4 Results
6.4.1 Geological observations on the volcanic field
The side-scan mapping using the TOBI system revealed a large area of highly reflective
lava flows between about 8◦ 45’S and 8◦ 51’S on the Mid-Atlantic Ridge (Fig. 6.2b). This
volcanic field is located on-axis within the rift valley. The rift valley itself is bounded by
normal faults which were identified by very high reflectivity, linear features on the TOBI
images (Fig. 6.2b,c). The highly reflective area is about 8 km long and has a maximum
width of 2 km. A map showing an interpretation of the TOBI image in terms of major
volcanic and tectonic structures is shown in Fig. 6.2c. The volcanic field surrounds two
rows of small volcanic cones, each of about 3 km length. These two elongated structures
probably formed above dikes and fed the surrounding lava flows. Several small faults occur
within the lava flows suggesting tectonic movement after the extrusion. Although the flow
is highly reflective in the side-scan sonar map, ROV observations show sediment patches
on the lavas (Fig. 6.3, 6.4). Consequently, the field did not form from recent volcanic
activity and no sign of hydrothermal activity was observed both during ROV tracks and
water sampling. The faulting may suggest that the segment evolves into a tectonic phase
following a magmatic stage similar to the observations on other parts of slow-spreading
ridges (Stakes et al., 1984). A pillow mound exists east of the southern row of small vol-
98
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
Figure 6.3: Summary of ROV dive 43, a W-E trending transect over the Southern lava unit and
located just south of the southern row of volcanic cones. A perspective image of combined bathymetry
and sidescan backscatter is shown to the top. A vertically exaggerated and illustrated crossection of the
transect with interpretation of structural features and lava flow morphology is shown in the middle. The
related geochemistry (K/Ti) over the transect can be seen to the bottom.
canic cones and this structure was studied by the ROV track along the field (ROV dive
44, station 159). This pillow mound is about 400 m in diameter and up to 50 m high. The
Dissertation P.A. Brandl
99
6. Compositional variation of lavas from a young volcanic field
mound consists entirely of pillow lavas and some single tubes. Lava patches were filled
by sediment and few Gorgonariae were observed. The southern end of one of the rows of
volcanic cones was traversed by ROV dive 43 (Station 155). Lava flow morphologies are
similar to the distinct pillow mound described before but tend to be lobate rather than
pillow lava.
Figure 6.4: Photographs of representative lavas from the ROV dives (for location of dives see
Fig. 6.2). (a) Typical pillow lavas with sediment pockets. (b) Lobate lava with single lava tubes. (c) Sheet
flow with thin sediment cover. (d) Lava pillar in collapse pit. (e) Aa-type lava overlying a lobate lava flow
(orange arrows). (f) Jumbled sheet flow overlying sedimented pillow lavas.
The lavas surrounding the volcanic cones show variable morphologies. Lavas along the
ROV track across the southern volcanic field (Fig. 6.2, 6.3) were dominated by pillow
lava with several centimetres of sediment in cavities (Fig. 6.3). Lobate and pancake lavas
100
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
were observed only in a few restricted areas. Figure 6.3 shows detailed observations (and
geochemical data, that will be discussed below) of ROV dive 43 (Station 155; see dive
track on Fig. 6.2 and 6.3). The dive track is located just south of the southern row of
volcanic cones in an area of high sidescan reflectivity (Fig. 6.3, top). At the western end,
pillow lavas are the predominant morphology but are lava flows are fractured (Fig. 6.3a,b)
and covered by thicker sediment than to the east. The southern tip of the row of volcanic
cones consist of pillow, lobate and pancake lava (Fig. 6.3c) with striated pillows on its
flanks (Fig. 6.3d). Further to the east, lava flow morphologies comprise pillow and lobate
lavas and minor sheet flows (Fig. 6.3e–g). Sediment cover and signs of tectonic activity
are increasing towards the eastern end of Dive 43 (Fig. 6.3h).
In contrast, lava flow morphologies along the field (ROV dive 44, station 159) are differing
in character. Except for the pillow mound (Fig. 6.4a), the volcanic field is dominated by
many distinct lava flows of different morphologies (Fig. 6.4). The area south of the pillow
mound is characterised by lobate lava (Fig. 6.4b) and sheet flows (and pancake lava; Fig.
6.4c,d) that are disrupted by distinct flows of pillow lava or jumbled sheet flows. At least
three different lava flows occur in this region. North of the pillow mound lava flow morphologies are dominated by lobate and pillow lava with minor sheet flows and jumbled
sheet flows (Fig. 6.4e,f). Camera observations indicate three different lava flows along the
ROV profile.
6.4.2 Petrography of the lavas
All lavas are fresh and have glassy rims with thin palagonite and Mn-Fe oxide staining.
Olivine and plagioclase are the most abundant phenocryst phases but clinopyroxene was
observed in a few samples (e.g., 159ROV-5). Phenocryst sizes range up to 10 mm for
plagioclase whereas the other minerals are generally smaller. Chromium-spinel occurs as
inclusions in olivine and clinopyroxene.
6.4.3 Composition of the volcanic glasses
The lavas sampled from the volcanic field and its surroundings show significant compositional variation with MgO contents ranging from 8.5 to 4.5 wt. % (Fig. 6.5). Three
different compositional groups of lavas can be distinguished in terms of K2 O and geographical occurrence. The largest group of samples has intermediate K2 O of 0.2 wt. %
at high MgO and shows a significant increase of K2 O with decreasing MgO (Fig. 6.5d).
These lavas are from the southern part of the young volcanic field and will be called the
southern lava unit. The group termed ‘northern lava unit’ has been sampled along the
northern row of volcanic cones and the four samples (160VSR-163VSR; Fig. 6.2a) are
Dissertation P.A. Brandl
101
6. Compositional variation of lavas from a young volcanic field
Figure 6.5: Major element diagrams of glass samples from the Mid-Atlantic Ridge axis at
8◦ 48’S: (a) TiO2 , (b) Al2 O3 , (c) CaO, (d) K2 O (all in wt. %), (e) Cl (ppm) and (f) S(ppm) versus
MgO (wt. %). Note that the older samples are much more depleted than lavas from the young volcanic
field (northern and southern lava unit) and that the northern lavas from the young flow are much more
enriched (higher TiO2 , Al2 O3 and K2 O at a given MgO) than the southern lavas. Cl versus MgO (e)
indicates assimilation of hydrothermally altered material by the relatively evolved melts.
distinguishable from the southern unit by higher TiO2 , K2 O, Al2 O3 and lower CaO (Fig.
6.5a–d). In contrast, old lavas sampled north of the young volcanic field (M62/5 156DS)
and from the more sedimented edges (155ROV-1 and 148VSR) have lower K2 O and CaO
contents and lower K/Ti ratios (Fig. 6.3, 6.6c) similar to the two samples from the large
pillow mound sampled with ROV dive 44 (Fig. 6.2, 6.5, 6.6c).
The glasses from the southern lava unit lie along a linear trend of decreasing CaO and
Al2 O3 between 4.4 and 8.5 wt. % MgO whereas the older lavas and the northern lava unit
102
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
samples tend to lower CaO but higher Al2 O3 (Fig. 6.5b,c). In terms of S we find that all
lavas show increasing S contents with decreasing MgO (Fig. 6.5d). The Cl concentrations
in most southern unit lavas remain constant between 6 and 8.4 wt. % MgO but increase
at lower MgO. Most older glasses have lower Cl contents than the southern unit samples
but lavas from the northern unit have higher Cl (Fig. 6.5e).
6.4.4 Along axis variations of compositions
Figure 6.6: Different geochemical parameters plotted versus latitude (◦ S): (a) MgO, (b)
(Ce/Yb)N , (c) K/Ti, (d) Cl/K, (e) 87 Sr/86 Sr and (f) (226 Ra/230 Th) shows the along axis variations
of the different lava units.
The MgO contents of the glasses seem to show systematic variations in the southern lava
unit with the highest MgO in the centre of this unit and lower MgO at the northern and
southern end (Fig. 6.6a). Both the northern lavas and the older glasses have relatively
Dissertation P.A. Brandl
103
6. Compositional variation of lavas from a young volcanic field
constant MgO. The compositional differences between the three units are observed in
K/Ti and primitive mantle normalised (Ce/Yb)N where the northern unit has the highest
K/Ti but intermediate (Ce/Yb)N whereas the older lavas are much more depleted than
the younger lava units (Fig. 6.6b,c). In terms of 87 Sr/86 Sr all lavas from the young lava
field are similar and have higher Sr isotope ratios than the older lavas (Fig. 6.6e). Most
glasses have similar Cl/K <0.07 but the more evolved glasses at the ends of the southern
unit indicate a larger variation and higher Cl/K ratios (Fig. 6.6d). In general, we observe
also different (226 Ra/230 Th) for the three different lava units of the MAR at 8◦ 48’S where
the northern lavas have the highest disequilibrium and most glasses from the southern
unit have (226 Ra/230 Th) of about 1.4 (Fig. 6.6f). The samples from the old lavas but also
three from the southern unit are in equilibrium. All lavas from the young volcanic field
have similar 87 Sr/86 Sr and 143 Nd/144 Nd but slight differences exist in (Ce/Yb)N (Fig. 6.7).
The older lavas are more depleted and have lower 87 Sr/86 Sr at similar 143 Nd/144 Nd.
Figure 6.7: Variation
of
(a)
(Ce/Yb)N and (b) 143 Nd/144 Nd
versus 87 Sr/86 Sr of the four different
lava units. Note that the lavas from
the young lava field (northern and
southern lava unit) have similar isotopic
compositions but differ in incompatible
element ratios, most notably in K/Ti
(Fig. 6.6c).
6.5 Discussion
6.5.1 Definition and formation of the lava flow units
Although the lavas in the 8◦ 48’S area show the same backscatter in the TOBI map
(Fig. 6.2b) the geochemical data indicate significant differences between the different
104
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
areas (Fig. 6.5, 6.6, 6.7). Based on the incompatible element compositions, two different
lava units with different magma sources can be defined in the young lavas and a third in
the older lavas surrounding the young eruptions. Interestingly, the solitary large pillow
mound occurring in the southern unit shows the same geochemical composition as the
older lava units on the flanks and to the north of the young flows (Fig. 6.5, 6.6, 6.7). It
thus appears that this pillow mound is older and was surrounded by the younger flow.
The southern unit is more extensive and occurs between 8◦ 51’S and 8◦ 47’S (i.e., over a
distance of 7 km) apparently overlapping with the northern unit. The sidescan reflectivity
does not reveal any difference between the two regions indicating very similar sediment
cover and age. The most voluminous lavas occur east of the southern row of volcanic cones
with a width of about 2 km and the whole southern unit covers an estimated area of 9
km2 . The northern unit surrounds the about 3 km long row of volcanic cones with a width
of perhaps 1.5 km, thus covering about 5 km2 . These two rows of small volcanic cones
probably respresent pillow mounds that formed above the feeder dykes of the eruptions
and are common on the Mid-Atlantic Ridge and have been termed ‘hummocks’ (Smith
and Cann, 1993; Yeo et al., 2012).
6.5.2 Composition and petrogenesis of the southern lava unit
We conclude that although the lava field formed during a short period of time it consists
of a variety of lavas suggesting both different parental magmas and variable degrees of
crystal fractionation. This implies that the eruption formed from several relatively small
batches of magma rising independently through the crust within a relatively short period of time. It also indicates that the mantle beneath this segment is heterogeneous on a
small scale of few kilometres. Incompatible element ratios and radiogenic isotopes indicate
a northward directed trend of decreasing enrichment in the lavas along the A2 segment
(Hoernle et al., 2011) but the lavas of the young field do not fit into that trend. Rather,
the northern lavas are more enriched than the southern lavas contrary to what would be
expected on the larger scale.
The glasses from the southern unit show a relatively large range of MgO between
8.4 and 4.4 wt. % implying that some of the melts experienced considerable amounts
of fractional crystallisation. Because MgO, CaO and Al2 O3 all decrease, the fractionated
phases must be olivine and plagioclase that are also the most abundant mineral phases
in the samples. The decreasing MgO contents at the northern and southern end of the
southern lava unit indicate that melts at the edges of the magma system stagnated for
longer periods of time in the crust than in the centre of the magma system. Interestingly,
we also observe higher Cl/K in the lavas at the two ends of the eruptive unit, implying
more assimilation of hydrothermally altered material during the ascent. We suggest that
Dissertation P.A. Brandl
105
6. Compositional variation of lavas from a young volcanic field
magma ascent in the southern unit was relatively fast in the centre of the dike but much
slower at the edges, leading to increased cooling, fractionation and assimilation at the two
tips of the dike.
6.5.3 Constraints on eruption ages
Figure 6.8: (226 Ra/230 Th) versus Ba/Th for the A2 lavas following the model of Rubin and Macdougall (1990). Based on Ba/Th and Ba concentrations we assume a mantle source with Ba/Th of about
20 and because all A2 lavas have similar Ba/Th we assume similar initial (226 Ra/230 Th). The lines indicate different ages assuming an initial (226 Ra/230 Th) of 4. A higher initial (226 Ra/230 Th) of 5 would
increase the ages by about 1000 a whereas a lower (226 Ra/230 Th) of 3 would imply ages of about 1000 a
less.
Short-lived isotopes like 230 Th and 226 Ra can be used to determine approximate ages of
lavas (Rubin and Macdougall, 1990; Sturm et al., 2000). The (226 Ra/230 Th) of the A2
lavas indicate at least three different volcanic events and in general, the different ages
correspond to the groups defined by geographical and incompatible element means (Fig.
6.6f). Most of the lavas from the lava field show significant Ra excesses indicating an age of
much less than 8 ka where the northern unit has the highest excesses (Fig. 6.6f). However,
there are three samples from the southern lava unit that are in equilibrium and thus must
be older than 8 ka. This implies that lavas with similar composition erupted over an
extensive period of time. However, most of the depleted lavas have (226 Ra/230 Th) of 1 and
thus must also been older than 8 ka. Rubin and Macdougall (1990) suggested that the
variation of (226 Ra/230 Th) relative to Ba/Th can be used to determine the ages of MORB.
We find that all lavas from the volcanic field have similar Ba/Th of 100 to 120 (Fig. 6.8)
106
Dissertation P.A. Brandl
6. Compositional variation of lavas from a young volcanic field
which suggests that they also should have had similar initial (226 Ra/230 Th) because Ra
behaves similarly to Ba during partial melting. According to this model the lavas from
the northern unit would be the youngest because they have the highest (226 Ra/230 Th).
Because the A2 MORB are more enriched than average depleted MORB their mantle
source probably has a higher Ba/Th of perhaps 20 compared to the estimated Ba/Th of 6
for depleted MORB (Rubin and Macdougall, 1990). If we assume an initial (226 Ra/230 Th)
of 4, the northern lavas would have ages of about 3 ka and the southern of about 4 ka
(Fig. 6.8). Although we cannot determine the initial (226 Ra/230 Th) we suggest it to be
4±1 (Rubin and Macdougall, 1990) so that these ages have an error of about 1 ka because
a lower initial ratio of 3 would lead to ages that are about 1 ka younger and a higher initial
(226 Ra/230 Th) of 5 would yield ages about 1 ka older. These ages are considerably younger
than the roughly 10 ka suggested for the Serocki volcano and axial volcanic ridge on the
northern MAR (Sturm et al., 2000). The different ages indicate that the volcanic activity
in segment A2 occurs over relatively brief periods of time following several thousand years
of tectonic activity only. At present, this part of the segment is in a tectonic stage as also
indicated by faults and abundance of sediments. The chemical differences between the
lavas of different age imply that magma sources on slow-spreading axes vary considerably
on timescales of several thousand years.
6.5.4 Magma ascent beneath the slow-spreading A2 segment
Segment A2 is close to the melting anomaly at 10◦ S on the MAR where crustal thickness reaches 14 km and where three off-axis seamounts exist (Minshull et al., 1998). The
crustal thickness in the area of the A2 lava field is about 10 km, implying an increased
magma production compared to average oceanic crust (Bruguier et al., 2003). The lava
flows erupted from two about 3 km long fissures that most likely represent dikes through
the uppermost 2 km of the crust. However, the geochemical differences indicate that the
magma reservoirs of the two dikes are separated also at greater depths and probably also
within the mantle. The smaller magma batch of the northern unit apparently ascended
later than the larger magma batch of the southern unit and the source of the former
became more enriched. The different magma sources for the southern and northern units
indicate that the magma transport occurs primarily vertical and lateral transport is restricted to perhaps 10 km. We conclude that there is only very small-scale lateral magma
transport in the slow-spreading crust at segment A2. This supports the model of single magma batches of small volume rising beneath slow-spreading ridges where magma
sources vary significantly in space and time.
Dissertation P.A. Brandl
107
6. Compositional variation of lavas from a young volcanic field
6.6 Conclusions
This study presents detailed geologic observations combined with geochemical data on
a young volcanic field located within the rift valley of the Southern Mid-Atlantic Ridge
(8◦ 48’S). So far, studies combining baythmetric and sidescan data with detailed observations and direct sampling by ROV are rare on slow-spreading ridges but provide important
insights into the formation of ocean crust at these ridge types. Our results from the magmatically robust segment A2 of the southern Mid-Atlantic Ridge (between Ascension
and Bode Verde Fracture Zone) have important implications for the interpretation of the
chemical composition of MORB, melt transport through the oceanic crust and the structure of the oceanic crust in general.
Our data indicate that the geochemical composition of MORB erupted at slow-spreading
ridges is variable not only in terms of chemical variation with time (few ka) but can also be
variable on a relatively small spatial scale (few km). This implies that magma underneath
the Mid-Atlantic Ridge (and probably other slow-spreading ridges) rise is small, chemically isolated batches that can erupt through distinct feeder dykes in close juxtaposition.
These feeder dykes are represented by elongated ridges of pillow mounds (‘hummocks’)
but the resulting lava flows have highly variable morphologies including sheet flows, lobate
lava and of course pillow lavas.
Lavas on the young volcanic field at 8◦ 48’S cover an age range of at least 5 ka. During that time the geochemical composition of lavas erupted changed significantly towards
a more enriched composition (e.g., higher K/Ti, 87 Sr/86 Sr). Youngest lavas (∼3-5 ka) are
represented by the northern lava unit also showing the most enriched chemical composition (e.g., K/Ti >0.4; 87 Sr/86 Sr ∼0.7025). This lava unit is located on the northward
dipping flank and furthest from the segment centre. In contrast, to common models, the
enriched composition of these lavas is not related to smaller degrees of partial melting as
seen in lower (Ce/Yb)N compared to the southern lava unit with an intermediate chemical composition. This implies that the mantle underneath this region of the southern
Mid-Atlantic Ridge is highly heterogeneous on small scales.
Acknowledgements
We thank Captain M. Kull and his crew for their help during cruise M64/1 with R/V
Meteor and the Bremen ROV team for their excellent work. We gratefully acknowledge the
help of P. Appel, B. Mader, and N. Stroncik with the electron beam microprobe analyses.
This work was funded by Deutsche Forschungsgemeinschaft under grants DE572/22-1 and
22-2 and HA2568/13-1.
108
Dissertation P.A. Brandl
7. Synthesis & Outlook
7 Synthesis & Outlook
7.1 The evolution of the upper mantle
Previous studies have argued for a globally hotter mantle potential temperature (by about
50◦ C) prior to 80 Ma (Humler et al., 1999) either as a consequence of a so-called mantle
avalanche (Machetel and Humler, 2003) or as a function of distance to continents and
heat transfer from below these to mid-ocean ridges (Humler and Besse, 2002).
In contrast, our study of ancient MORB (chapter 3) shows that different ocean basins
followed different thermal evolution paths. The existence of a supercontinent significantly
influences the thermal structure of the upper mantle as suggested by theoretical models
(e.g., Coltice et al., 2012; Phillips and Coltice, 2010; Rolf et al., 2012). A single and very
large coherent landmass, such as Pangaea through the Permian and Triasssic, works very
efficiently as an insulating lid above the underlying mantle. This insulation effect prevents
the heat produced in the Earth’s interior from escaping to the surface by convection. Since
conductive heat transfer is much less efficient than convective transfer, the heat is retained
underneath a supercontinental landmass resulting in a significant increase in mantle potential temperature.
Our data support the idea of continental insulation by indicating very high mantle potential temperatures recorded in MORB erupted immediately after the breakup of Pangaea
and the opening of the Central Atlantic in the Jurassic. In contrast, Pacific MORB show
no clear systematic variation with time and the overall magnitude of change in mantle
potential temperature is much smaller. The Pacific as an ocean (not the Pacific plate
itself) exists for hundreds of million years, so heat transfer from the suboceanic mantle to
the surface by mantle convection is present for extremely long periods of time resulting
in none or only minor effects by continental insulation underneath Pangaea. Literature
data from the Indian Ocean (and the present situation along the mid-ocean ridges in the
Red Sea, the Gulf of Aden and along the Central Indian Ridge to the Southwest Indian
Ridge) indicate a thermal evolution similar to the Atlantic.
The study presented in chapter 3, is probably the first to show convincingly the effects of
Dissertation P.A. Brandl
109
7. Synthesis & Outlook
continental insulation on mantle potential temperature and thus on the geochemistry of
MORB. Unfortunately, the number of suitable samples is very limited since ocean drilling
is expensive and time-consuming. Fresh glasses from old seafloor are rare and my samples might still be not fully representative for the complete ocean basins. Future detailed
studies of old seafloor combining geophysical and geochemical methods would thus help
to improve our understanding of mantle evolution and mantle convection through the
Earth’s history.
7.2 Constraints on mantle melting and mixing processes
In chapter 3, I have shown that the major element composition of MORB does faithfully record changes in the degree of partial melting and thus also in mantle potential
temperature as a result of continental insulation. Responsible for the preservation of the
record of mantle potential temperature is the overall large degree of partial melting and
the effective mixing processes during the petrogenesis of MORB.
However, the study on post-spreading volcanism at the Galapagos Rise presented in chapter 5 demonstrates how the chemical composition of MORBs change during the extinction of spreading and decreasing melting degrees. During an early stage of post-spreading
volcanism, erupted lavas still have a composition similar to normal depleted N-MORB
whereas during a later stage, when the overall degree of partial melting is significantly
lower, the erupted basalts evolve towards more enriched compositions.
This effect is even better observed in rocks from Seamount 6 (chapter 4). The major
and trace element and radiogenic isotope data of Seamount 6 basalts provide convincing
evidence for a two-component melting process. We used state-of-the-art melting models
(e.g., Stracke and Bourdon, 2009; Ito and Mahoney, 2005a,b; Ingle et al., 2010) to further constrain the dynamics of melting in the mantle. The implications of this study are:
a) different mantle rocks do not contribute equally to the accumulated melt, b) the quantity of enriched material in the mantle is on the order of 5–10%, c) the more fertile
lithology (with a lower solidus) becomes progressively more diluted in the melt derived
from the ambient depleted mantle peridotite with increasing degrees of partial melting
(with increasing proportions of melt from the shallow and depleted mantle) and d) the
chemical composition of the mantle source might be more extreme than preserved in the
erupted lavas.
To summarise, it is very likely that the mantle is heterogeneous even underneath midocean ridges. The heterogeneity is a function of the overall degree of partial melting and
110
Dissertation P.A. Brandl
7. Synthesis & Outlook
the effectiveness of mixing during the process of melt aggregation that controls whether the
source heterogeneity would still be preserved in the chemical composition of the erupted
lava. Small degrees of partial melting and a very narrow magma plumbing system (e.g.,
ocean islands or off-axial seamounts) will preserve and record the chemical composition
of the source more faithfully. In contrast, MORB are generated through large degrees of
partial melting and the primary melts are aggregated in long-living magma chambers that
are efficient in mixing and homogenisation of melt. The chemical composition of MORB
will thus reflect degree of partial melting, whereas basalts generated by low degrees of
partial melting will best record information of mantle heterogeneity and the melt extraction processes.
The best potential to further investigate mantle source heterogeneity and melting and
mixing processes may be provided by studying melt inclusions trapped in minerals and
brought to the surface without (or only minor) chemical equilibration. Recently developed
analytical techniques such as ionprobe techniques, fourier-transform infrared spectrometry or synchrotron techniques will allow further constrains on melting processes and redox
conditions in the mantle.
7.3 Volcanic eruptions at mid-ocean ridges
The study of a young volcanic field presented in chapter 6 allows to infer on magmatic
processes at slow-spreading ridges. Data from this study indicate that the geochemical
composition of MORB at these ridges is variable with time (within a few thousand years)
but also on a relatively small spatial scale of only a few kilometres. This implies that
magma underneath slow-spreading ridges (at least at magmatically robust segments) rise
in small, chemically isolated batches that are erupted in close juxtaposition by feeder dyke
eruptions. The elongated rows of small volcanic cones (‘hummocks’) most likely represent
such feeder dykes and are mainly composed of pillow lava. In contrast, surrounding lava
flows have highly variable morphologies including (jumbled) sheet flows and lobate and
pancake lavas.
Further implications of this study give evidence a) for a chemically heterogeneous mantle underneath this region of the southern Mid-Atlantic Ridge and b) that the MORB
erupted at slow-spreading ridges are fed by small, chemically isolated batches of magma.
This would imply that the extrusive layer of oceanic crust formed at slow-spreading ridges
might be morphologically and chemically highly heterogeneous.
Further studies combining visual observations, detailed mapping and sampling and in-
Dissertation P.A. Brandl
111
7. Synthesis & Outlook
tense geochemical studies are required to put further constraints on the plumbing systems
at slow-spreading ridges and to investigate the process that form the upper oceanic crust.
Thus, it would be useful to re-visit previously studied locations at slow-spreading ridges
and to intensify mapping and chemical studies of distinct lava flows.
112
Dissertation P.A. Brandl
References
References
Abbott, D., Burgess, L., Longhi, J., and Smith, W. H. F. (1994). An empirical thermal history of the
Earth’s upper mantle. Journal of Geophysical Research, 99(B7):13835–13850.
Abelson, M., Baer, G., and Agnon, A. (2001). Evidence from gabbro of the Troodos ophiolite for lateral
magma transport along a slow-spreading mid-ocean ridge. Nature, 409(6816):72–75.
Abouchami, W., Galer, S. J. G., and Hofmann, A. W. (2000). High precision lead isotope systematics of
lavas from the Hawaiian Scientific Drilling Project. Chemical Geology, 169(1-2):187–209.
Aggrey, K. E., Muenow, D. W., and Batiza, R. (1988). Volatile abundances in basaltic glasses from
seamounts flanking the East Pacific Rise at 21 ◦ N and 12–14 ◦ N. Geochimica et Cosmochimica Acta,
52(8):2115–2119.
Albarède, F. (1995). Introduction to Geochemical Modeling. Cambridge University Press, Cambridge.
Allan, J. F., Batiza, R., Perfit, M. R., Fornari, D. J., and Sack, R. O. (1989). Petrology of lavas from the
Lamont seamount chain and adjacent East Pacific Rise, 10 ◦ N. Journal of Petrology, 30(5):1245–1298.
Allègre, C. J. (1982). Chemical geodynamics. Tectonophysics, 81(3):109–132.
Allègre, C. J., Brevart, O., Dupré, B., and Minster, J. F. (1980). Isotopic and chemical effects produced
in a continuously differentiating convecting Earth mantle. Philosophical Transactions of the Royal
Society A: Mathematical, Physical and Engineering Sciences, 297(1431):447–477.
Allègre, C. J. and Minster, J. F. (1978). Quantitative models of trace element behavior in magmatic
processes. Earth and Planetary Science Letters, 38(1):1–25.
Allègre, C. J. and Turcotte, D. L. (1986). Implications of a two-component marble-cake mantle. Nature,
323(6084):123–127.
Allen, R. M., Nolet, G., Morgan, W. J., Vogfjörd, K., Bergsson, B. H., Elendsson, P., Foulger, G. R.,
Jakobsdóttir, S., Julian, B. R., Pritchard, M., Ragnarsson, S., and Stefánsson, R. (2002). Imaging the
mantle beneath Iceland using integrated seismological techniques. Journal of Geophysical Research,
107(B12):2325.
Anderson, D. L. (1982).
297(5865):391–393.
Hotspots, polar wander, Mesozoic convection and the geoid.
Nature,
Anderson, R. N. and Sclater, J. G. (1972). Topography and evolution of the East Pacific Rise between
5 ◦ S and 20 ◦ S. Chemical Geology, 14(3):433–441.
Arevalo Jr., R. and McDonough, W. F. (2010). Chemical variations and regional diversity observed in
MORB. Chemical Geology, 271(1-2):70–85.
Armstrong, R. L. (1968). A model for the evolution of strontium and lead isotopes in a dynamic Earth.
Reviews of Geophysics, 6(2):175–199.
Asimow, P. D., Dixon, J. E., and Langmuir, C. H. (2004). A hydrous melting and fractionation model
for mid-ocean ridge basalts: Application to the Mid-Atlantic Ridge near the Azores. Geochemistry
Geophysics Geosystems, 5(1):Q01E16.
Asimow, P. D. and Langmuir, C. H. (2003). The importance of water to oceanic mantle melting regimes.
Nature, 421(6925):815–820.
Batiza, R. (1977). Petrology and chemistry of Guadalupe Island: An alkalic seamount on a fossil ridge
crest. Geology, 5(12):760–764.
Batiza, R. (1980). Origin and petrology of young oceanic central volcanoes: Are most tholeiitic rathern
than alkalic? Geology, 8(10):477–482.
Batiza, R. (1989). Failed rifts. In Winterer, E. L., Hussong, D. M., and Decker, R. W., editors, The
Dissertation P.A. Brandl
113
References
Eastern Pacific Ocean and Hawaii, pages 177–186. The Geology of North America, Boulder.
Batiza, R. and Chase, C. G. (1981).
289(5800):787–789.
Miocene spreading centre south of Isla Guadalupe.
Nature,
Batiza, R., Oestrike, R., and Futa, K. (1982). Chemical and isotopic diversity in basalts dredged from
the East Pacific Rise at 10 ◦ S, the fossil Galapagos Rise and the Nazca plate. Marine Geology, 49(12):115–132.
Batiza, R., Smith, T., and Niu, Y. (1989). Geological and petrologic evolution of seamounts near the
EPR based on submersible and camera study. Marine Geophysical Researches, 11(3):169–236.
Batiza, R. and Vanko, D. (1983). Volcanic development of small oceanic central volcanoes on the flanks of
the East Pacific Rise inferred from narrow-beam echo-sounder surveys. Marine Geology, 54(1-2):53–90.
Batiza, R. and Vanko, D. (1984). Petrology of young Pacific seamounts. Journal of Geophysical Research,
89(B13):11235–11260.
Batiza, R. and Vanko, D. (1985). Petrologic evolution of large failed rifts in the eastern Pacific: Petrology
of volcanic and plutonic rocks from the Mathematician Ridge area and the Guadalupe Trough. Journal
of Petrology, 26(3):564–602.
Batiza, R. and Watts, A. (1986). Scientific seamount drilling - Report of a JOI-USSAC sponsored
workshop. Technical report, Lamont-Doherty Geological Observatory.
Beier, C., Haase, K. M., and Turner, S. P. (2012a). Conditions of melting beneath the Azores. Lithos,
144-145:1–11.
Beier, C., Mata, J., Stöckhert, F., Mattielli, N., Brandl, P. A., Madureira, P., Genske, F. S., Martins, S.,
Madeira, J., and Haase, K. M. (2012b). Geochemical evidence for melting of carbonated peridotite on
Santa Maria Island, Azores. Contributions to Mineralogy and Petrology. in press.
Beier, C., Vanderkluysen, L., Regelous, M., Mahoney, J. J., and Garbe-Schönberg, D. (2011). Lithospheric
control on geochemical composition along the Louisville Seamount Chain. Geochemistry Geophysics
Geosystems, 12(9):Q0AM01.
Berner, R. A. and Kothavala, Z. (2001). GEOCARB III: A revised model of atmospheric CO2 over
Phanerozoic time. American Journal of Science, 301:182–204.
Bienvenu, P., Bougault, H., Joron, J. L., Treuil, M., and Dmitriev, L. (1990). MORB alteration: Rareearth element/non-rare-earth hygromagmaphile element fractionation. Chemical Geology, 82:1–14.
Bodinier, J. I., Merlet, C., Bedini, R. M., Simien, F., Remaidi, M., and Garrido, C. J. (1996). Distribution
of niobium, tantalum, and other highly incompatible trace elements in the lithospheric mantle: The
spinel paradox. Geochimica et Cosmochimica Acta, 60(3):545–550.
Bohrson, W. A. and Reid, M. R. (1995). Petrogenesis of alkaline basalts from Socorro Island, Mexico:
Trace element evidence for contamination of ocean island basalt in the shallow ocean crust. Journal
of Geophysical Research, 100(B12):24555–24576.
Bottinga, Y. and Allègre, C. J. (1973). Thermal aspects of sea-floor spreading and the nature of the
oceanic crust. Chemical Geology, 18(1-2):1–17.
Boudier, F. and Nicolas, A. (1985). Harzburgite and lherzolite subtypes in ophiolitic and oceanic environments. Earth and Planetary Science Letters, 76(1-2):84–92.
Bourdon, B., Zindler, A., Elliott, T., and Langmuir, C. H. (1996). Constraints on mantle melting at
mid-ocean ridges from global 238 U–230 Th disequilibrium data. Nature, 384:231–235.
Bowen, N. L. (1928). The Evolution of the Igneous Rocks. Princeton University Press.
Bown, J. W. and White, R. S. (1994). Variation with spreading rate of oceanic crustal thickness and
geochemistry. Earth and Planetary Science Letters, 121(3-4):435–449.
Brandl, P. A., Beier, C., Regelous, M., Abouchami, W., Haase, K. M., Garbe-Schönberg, D., and Galer,
S. J. G. (2012). Volcanism on the flanks of the East Pacific Rise: Quantitative constraints on mantle
heterogeneity and melting processes. Chemical Geology, 298-299:41–56.
Bruguier, N. J., Minshull, T. A., and Brozena, J. M. (2003). Morphology and tectonics of the Mid-Atlantic
Ridge, 7 ◦ –12 ◦ S. Journal of Geophysical Research, 108(B2):2093.
114
Dissertation P.A. Brandl
References
Bryan, W. B. (1983). Systematics of modal phenocryst assemblages in submarine basalts: Petrologic
implications. Contributions to Mineralogy and Petrology, 83(1-2):62–74.
Buck, W. R., Carbotte, S. M., and Mutter, C. (1997). Controls on extrusion at mid-ocean ridges. Geology,
25:935–938.
Buck, W. R. and Su, W. (1989). Focused mantle upwelling below mid-ocean ridges due to feedback
between viscosity and melting. Geophysical Research Letters, 16(7):641–644.
Calcagno, P. and Cazenave, A. (1994). Subsidence of the seafloor in the Atlantic and Pacific Oceans:
Regional and large-scale variations. Earth and Planetary Science Letters, 126(4):473–492.
Campbell, I. H. and Griffiths, R. W. (1990). Implications of mantle plume structure for the evolution of
flood basalts. Earth and Planetary Science Letters, 99(1):79–93.
Campbell, I. H. and Jarvis, G. T. (1984). Mantle convection and early crustal evolution. Chemical
Geology, 26(1):15–56.
Canales, J. P., Detrick, R. S., Carbotte, S. M., Kent, G. M., Diebold, J. B., Harding, A., Babcock,
J., Nedimović, M. R., and van Ark, E. (2005). Upper crustal structure and axial topography at
intermediate spreading ridges: Seismic constraints from the southern Juan de Fuca Ridge. Journal of
Geophysical Research, 110:B12104.
Cann, J. R. (1968). Geological Processes at Mid-Ocean Ridge Crests. Geophysical Journal International,
15(3):331–341.
Cann, J. R. (1974). A Model for Oceanic Crystal Structure Developed. Geophysical Journal International,
39(1):169–187.
Carbotte, S., Mutter, C., Mutter, J., and Ponce-Correa, G. (1998). Influence of magma supply and
spreading rate on crustal magma bodies and emplacement of the extrusive layer: Insights from the
East Pacific Rise at lat 16 ◦ N. Geology, 26(5):455–458.
Castillo, P. R., Clague, D. A., Davis, A. S., and Lonsdale, P. F. (2010). Petrogenesis of Davidson Seamount
lavas and its implications for fossil spreading center and intraplate magmatism in the eastern Pacific.
Geochemistry Geophysics Geosystems, 11(2):Q02005.
Castillo, P. R., Klein, E., Bender, J., Langmuir, C. H., Shirey, S., Batiza, R., and White, W. (2000).
Petrology and Sr, Nd, and Pb isotope geochemistry of mid-ocean ridge basalt glasses from the 11 ◦ 45’N
to 15 ◦ 00’N segment of the East Pacific Rise. Geochemistry Geophysics Geosystems, 1:1999GC000024.
Chase, C. G. (1981). Oceanic island Pb: Two-stage histories and mantle evolution. Chemical Geology,
52(2):277–284.
Choe, W. H., Lee, J. I., Lee, M. J., Hur, S. D., and Jin, Y. K. (2007). Origin of E-MORB in a fossil
spreading center: The Antarctic-Phoenix Ridge, Drake Passage, Antarctica. Geosciences Journal,
11(3):185–199.
Choi, S. H., Choe, W. H., and Lee, J. I. (2007). Mantle heterogeneity beneath the Antarctic-Phoenix
Ridge off Antarctic Peninsula. Island Arc, 17(1):172–182.
Choukroune, P., Francheteau, J., and Hékinian, R. (1984). Tectonics of the East Pacific Rise near 12 ◦ 50’N:
A submersible study. Earth and Planetary Science Letters, 68(1):115–127.
Cipriani, A., Brueckner, H. K., Bonatti, E., and Brunelli, D. (2004). Oceanic crust generated by elusive
parents: Sr and Nd isotopes in basalt-peridotite pairs from the Mid-Atlantic Ridge. Geology, 32(8):657–
660.
Clague, D. A., Paduan, J. B., Duncan, R. A., Huard, J. J., Davis, A. S., Castillo, P. R., Lonsdale, P., and
DeVogelaere, A. (2009). Five million years of compositionally diverse, episodic volcanism: Construction
of Davidson Seamount atop an abandoned spreading center. Geochemistry Geophysics Geosystems,
10(12):Q12009.
Cogné, J. P., Humler, E., and Courtillot, V. (2006). Mean age of oceanic lithosphere drives eustatic
sea-level change since Pangea breakup. Earth and Planetary Science Letters, 245(1-2):115–122.
Coltice, N., Bertrand, H., Rey, P., Jourdan, F., Phillips, B. R., and Ricard, Y. (2009). Global warming
of the mantle beneath continents back to the Archaean. Gondwana Research, 15(3-4):254–266.
Dissertation P.A. Brandl
115
References
Coltice, N., Phillips, B. R., Bertrand, H., Ricard, Y., and Rey, P. (2007). Global warming of the mantle
at the origin of flood basalts over supercontinents. Geology, 35(5):391.
Coltice, N., Rolf, T., Tackley, P. J., and Labrosse, S. (2012). Dynamic causes of the relation between area
and age of the ocean floor. Science, 336(6079):336–338.
Condie, K. C. (2000). Episodic continental growth models: Afterthoughts and extensions. Tectonophysics,
322(1-2):153–162.
Courtillot, V., Besse, J., Vandamme, D., Montigny, R., Jaeger, J. J., and Cappetta, H. (1986). Deccan
flood basalts at the Cretaceous/Tertiary boundary? Earth and Planetary Science Letters, 80(3):361–
374.
Cox, K. G. (1989). The role of mantle plumes in the development of continental drainage patterns.
Nature, 342:873–877.
Crisp, J. A. (1984). Rates of magma emplacement and volcanic output. Journal of Volcanology and
Geothermal Research, 20:177–211.
Dalrymple, G. B. (2001). The age of the Earth in the twentieth century: A problem (mostly) solved.
Geological Society London Special Publications, 190(1):205–221.
Dasgupta, R. and Hirschmann, M. M. (2006). Melting in the Earth’s deep upper mantle caused by carbon
dioxide. Nature, 440(7084):659–662.
Dasgupta, R., Hirschmann, M. M., and Withers, A. C. (2004). Deep global cycling of carbon constrained
by the solidus of anhydrous, carbonated eclogite under upper mantle conditions. Earth and Planetary
Science Letters, 227(1):73–85.
Davis, A. S., Clague, D. A., Bohrson, W. A., Dalrymple, G. B., and Greene, H. G. (2002). Seamounts
at the continental margin of California: A different kind of oceanic intraplate volcanism. Geological
Society of America Bulletin, 114(3):316–333.
Davis, A. S., Gunn, S. H., Bohrson, W. A., Gray, L. B., and Hein, J. R. (1995). Chemically diverse,
sporadic volcanism at seamounts offshore southern and Baja California. Geological Society of America
Bulletin, 107(5):554–570.
DeMets, C., Gordon, R. G., Angus, D. F., and Stein, S. (1994). Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophysical Research Letters,
21:2191–2194.
DePaolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional
crystallization. Earth and Planetary Science Letters, 53(2):189–202.
DePaolo, D. J. and Manga, M. (2003). Deep origin of hotspots – the mantle plume model. Science,
300(5621):920–921.
DePaolo, D. J. and Wasserburg, G. J. (1976). Inferences about magma sources and mantle structure
from variations of 143 Nd/144 Nd. Geophysical Research Letters, 3(12):743–746.
Devey, C. W., Lackschewitz, K., and Baker, E. (2005). Hydrothermal and volcanic activity found on the
southern Mid-Atlantic Ridge. Eos, Transactions, AGU, 86(22):209 & 212.
Donnelly, K. E., Goldstein, S. L., Langmuir, C. H., and Spiegelman, M. (2004). Origin of enriched ocean
ridge basalts and implications for mantle dynamics. Earth and Planetary Science Letters, 226(3-4):347–
366.
Duncan, R. A. and Keller, R. A. (2004). Radiometric ages for basement rocks from the Emperor
Seamounts, ODP Leg 197. Geochemistry Geophysics Geosystems, 5(8):Q08L03.
Dungan, M. A. and Rhodes, J. M. (1978). Residual glasses and melt inclusions in basalts from DSDP Legs
45 and 46: Evidence for magma mixing. Contributions to Mineralogy and Petrology, 67(4):417–431.
Eakins, B. W. and Lonsdale, P. F. (2003). Structural patterns and tectonic history of the Bauer microplate, eastern tropical Pacific. Marine Geophysical Researches, 24(3-4):171–205.
Eason, D. and Sinton, J. (2006). Origin of high-Al N-MORB by fractional crystallization in the upper
mantle beneath the Galápagos Spreading Center. Earth and Planetary Science Letters, 252:423–436.
Elderfield, H. and Schultz, A. (1996). Mid-ocean ridge hydrothermal fluxes and the chemical composition
116
Dissertation P.A. Brandl
References
of the ocean. Annual Review of Earth and Planetary Sciences, 24(1):191–224.
Elkins, L. J., Gaetani, G. A., and Sims, K. W. W. (2008). Partitioning of U and Th during garnet
pyroxenite partial melting: Constraints on the source of alkaline ocean island basalts. Earth and
Planetary Science Letters, 265(1-2):270–286.
Ellam, R. M. (1992). Lithospheric thickness as a control on basalt geochemistry. Geology, 20(2):153–156.
Engel, A. E. J. and Engel, C. G. (1964a). Composition of basalts from the Mid-Atlantic Ridge. Science,
144(3624):1330–1333.
Engel, A. E. J. and Engel, C. G. (1964b). Igneous Rocks of the East Pacific Rise: The alkali volcanic suite
appear to be differentiated from a tholeiitic basalt extruded from the mantle. Science, 146(3643):477.
Engel, C. G., Fischer, R. L., and Engel, A. E. J. (1965). Igneous rocks of the Indian Ocean floor. Science,
150(3696):605–610.
Falloon, T. J. and Green, D. H. (1987). Anhydrous partial melting of MORB pyrolite and other peridotite
compositions at 10 kbar: Implications for the origin of primitive MORB glasses. Contributions to
Mineralogy and Petrology, 37(3-4):181–219.
Falloon, T. J. and Green, D. H. (1988). Anhydrous partial melting of peridotite from 8 to 35 kb and the
petrogenesis of MORB. Journal of Petrology, Special Volume 1:379–414.
Fisk, M. and Kelley, K. A. (2002). Probing the Pacific’s oldest MORB glass: Mantle chemistry and melting
conditions during the birth of the Pacific Plate. Earth and Planetary Science Letters, 202(3-4):741–752.
Fontignie, D. and Schilling, J. G. (1996). Mantle heterogeneities beneath the South Atlantic: A Nd-SrPb isotope study along the Mid-Atlantic Ridge (3 ◦ S–46 ◦ S). Earth and Planetary Science Letters,
142(1-2):209–221.
Forsyth, D. W. (1977). The evolution of the upper mantle beneath mid-ocean ridges. Chemical Geology,
38(1-2):89–118.
Fyfe, W. S. (1978). The evolution of the Earth’s crust: Modern plate tectonics to ancient hot spot
tectonics? Chemical Geology, 23(1-4):89–114.
Gaffin, S. (1987). Ridge volume dependence on seafloor generation rate and inversion using long term
sealevel change. American Journal of Science, 287(6):596.
Gale, N. H. and Mussett, A. E. (1973). Episodic uranium-lead models and the interpretation of variations
in the isotopic composition of lead in rocks. Reviews of Geophysics, 11(1):37–86.
Galer, S. J. G. (1999). Optimal double and triple spiking for high precision lead isotopic measurement.
Chemical Geology, 157(3-4):255–274.
Galer, S. J. G. and Abouchami, W. (1998). Practical application of lead triple spiking for correction of
instrumental mass discrimination. Mineralogical Magazine, (62A):491–492.
Garbe-Schönberg, D. (1993). Simultaneous determination of thirty-seven trace elements in twenty-eight
international rock standards by ICP-MS. Geostandards Newsletter, 17(1):81–97.
Gast, P. W., Tilton, G. R., and Hedge, C. (1964). Isotopic composition of lead and strontium from
Ascension and Gough Islands. Science, 145:1181–1185.
Govindaraju, K. (1995). 1995 working values with confidence limits for twenty-six CRPD, ANRT and
IW-GIT geostandards. Geostandards Newsletter, 19:1–32.
Graham, D. W., Jenkins, W. J., Kurz, M. D., and Batiza, R. (1987). Helium isotope disequilibrium and
geochronology of glassy submarine basalts. Nature, 326:384–386.
Graham, D. W., Zindler, A., Kurz, M. D., Jenkins, W. J., Batiza, R., and Staudigel, H. (1988). He,
Pb, Sr and Nd isotope constraints on magma genesis and mantle heterogeneity beneath young Pacific
seamounts. Contributions to Mineralogy and Petrology, 99(4):446–463.
Green, D. H. and Liebermann, R. C. (1976). Phase equilibria and elastic properties of a pyrolite model
for the oceanic upper mantle. Chemical Geology, 32(1-2):61–92.
Green, D. H. and Ringwood, A. E. (1967). The genesis of basaltic magmas. Contributions to Mineralogy
and Petrology, 15(2):103–190.
Dissertation P.A. Brandl
117
References
Grigné, C. and Labrosse, S. (2001). Effects of continents on Earth cooling: Thermal blanketing and
depletion in radioactive elements. Geophysical Research Letters, 28(14):2707–2710.
Haase, K. M. (1996). The relationship between the age of the lithosphere and the composition of oceanic
magmas: Constraints on partial melting, mantle sources and the thermal structure of the plates. Earth
and Planetary Science Letters, 144(1-2):75–92.
Haase, K. M., Beier, C., Fretzdorff, S., Leat, P. T., Livermore, R. A., Barry, T. L., Pearce, J. A., and
Hauff, F. (2011a). Magmatic evolution of a dying spreading axis: Evidence for the interaction of
tectonics and mantle heterogeneity from the fossil Phoenix Ridge, Drake Passage. Chemical Geology,
280(1-2):115–125.
Haase, K. M., Regelous, M., Duncan, R. A., Brandl, P. A., Stroncik, N., and Grevemeyer, I. (2011b).
Insights into mantle composition and mantle melting beneath mid-ocean ridges from postspreading
volcanism on the fossil Galapagos Rise. Geochemistry Geophysics Geosystems, 12(5):Q0AC11.
Haase, K. M. and Shipboard Scientific Party (2002). GARIMAG-Magmatism at the fossil spreading axis
of the Galapagos Rise, SE Pacific. Technical report, Institut für Geowissenschaften der ChristianAlbrechts-Universität Kiel.
Hall, L. S. and Sinton, J. M. (1996). Geochemical diversity of the large lava field on the flank of the East
Pacific Rise at 8 ◦ 17’S. Earth and Planetary Science Letters, 142(1-2):241–251.
Hammouda, T. (2003). High-pressure melting of carbonated eclogite and experimental constraints on
carbon recycling and storage in the mantle. Chemical Geology, 214(1-2):357–368.
Hanan, B. B., Kingsley, R. H., and Schilling, J. G. (1986). Pb isotope evidence in the South Atlantic for
migrating ridge–hotspot interactions. Nature, 322:137–144.
Hardie, L. A. (1996). Secular variation in seawater chemistry: An explanation for the coupled secular
variation in the mineralogies of marine limestones and potash evaporites over the past 600 m.y. Geology,
24(3):279–283.
Hart, S. R. (1984). A large-scale isotope anomaly in the Southern Hemisphere mantle.
309(5971):753–757.
Nature,
Hart, S. R. (1993). Equilibration during mantle melting: A fractal tree model. Proceedings of the National
Academy of Sciences, 90(24):11914–11918.
Harvey, J., Gannoun, A., Burton, K. W., Rogers, N. W., Alard, O., and Parkinson, I. J. (2006). Ancient
melt extraction from the oceanic upper mantle revealed by Re-Os isotopes in abyssal peridotites from
the Mid-Atlantic ridge. Earth and Planetary Science Letters, 244(3-4):606–621.
Hayes, D. E. (1988). Age-depth relationships and depth anomalies in the Southeast Indian Ocean and
South Atlantic Ocean. Journal of Geophysical Research, 93(B4):2937–2954.
Hébert, H., Villemant, B., Deplus, C., and Diament, M. (1999). Contrasting geophysical and geochemical
signatures of a volcano at the axis of the Wharton Fossil Ridge (N-E Indian Ocean). Geophysical
Research Letters, 26(8):1053.
Herron, E. M. (1972). Sea-Floor Spreading and the Cenozoic History of the East-Central Pacific. Geological Society of America Bulletin, 83:1671–1692.
Hertogen, J., Janssens, M., and Palme, H. (1980). Trace elements in ocean ridge basalt glasses: Implications for fractionations during mantle evolution and petrogenesis. Geochimica et Cosmochimica Acta,
44(12):2125–2143.
Herzberg, C. (2004). Partial crystallization of mid-ocean ridge basalts in the crust and mantle. Journal
of Petrology, 45(12):2389–2405.
Herzberg, C. and Asimow, P. D. (2008). Petrology of some oceanic island basalts: PRIMELT2.XLS
software for primary magma calculation. Geochemistry Geophysics Geosystems, 9(9):Q09001.
Herzberg, C., Asimow, P. D., Arndt, N., Niu, Y., Lesher, C. M., Fitton, J. G., Cheadle, M. J., and
Saunders, A. D. (2007). Temperatures in ambient mantle and plumes: Constraints from basalts,
picrites, and komatiites. Geochemistry Geophysics Geosystems, 8(2):Q02006.
Herzberg, C. and O’Hara, M. J. (2002). Plume-associated ultramafic magmas of Phanerozoic age. Journal
of Petrology, 43(10):1857–1883.
118
Dissertation P.A. Brandl
References
Hirano, N., Takahashi, E., Yamamoto, J., Abe, N., Ingle, S. P., Kaneoka, I., Hirata, T., Kimura, J. I.,
Ishii, T., Ogawa, Y., Machida, S., and Suyehiro, K. (2006). Volcanism in response to plate flexure.
Science, 313(5792):1426–1428.
Hirschmann, M. M. and Stolper, E. M. (1996). A possible role for garnet pyroxenite in the origin of the
‘garnet signature’ in MORB. Contributions to Mineralogy and Petrology, 124(2):185–208.
Hoernle, K., Hauff, F., Kokfelt, T. F., Haase, K. M., Garbe-Schönberg, D., and Werner, R. (2011). Onand off-axis chemical heterogeneities along the South Atlantic Mid-Ocean-Ridge (5–11 ◦ S): Shallow
or deep recycling of ocean crust and/or intraplate volcanism? Earth and Planetary Science Letters,
306(1-2):86–97.
Hoernle, K. and Tilton, G. R. (1991). Sr–Nd–Pb isotope data for Fuerteventura (Canary Islands) basal
complex and subaerial volcanics: Application to magma genesis and evolution. Schweizerische Mineralogische und Petrographische Mitteilungen, 71:3–18.
Hofmann, A. W. (1988). Chemical differentiation of the Earth: The relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters, 90(3):297–314.
Hofmann, A. W. (1997).
385(6613):219–229.
Mantle geochemistry: The message from oceanic volcanism.
Nature,
Hofmann, A. W. (2003). Sampling mantle heterogeneity through oceanic basalts: Isotopes and trace
elements. In Holland, H. D. and Turekian, K. K., editors, Treatise on geochemistry, pages 1–44.
Elsevier, Amsterdam.
Hofmann, A. W. and White, W. M. (1982). Mantle plumes from ancient oceanic crust. Earth and
Planetary Science Letters, 57:421–436.
Honda, M., Bernatowicz, T., Podosek, F. A., Batiza, R., and Taylor, P. T. (1987). Age determinations of
eastern Pacific seamounts (Henderson, 6 and 7)–Implications for near-ridge and intraplate volcanism.
Marine Geology, 74(1-2):79–84.
Hooft, E. E. E., Schouten, H., and Detrick, R. S. (1996). Constraining crustal emplacement processes
from the variation in seismic layer 2A thickness at the East Pacific Rise. Earth and Planetary Science
Letters, 142:289–309.
Humler, E. and Besse, J. (2002). A correlation between mid-ocean-ridge basalt chemistry and distance
to continents. Nature, 419(6907):607–609.
Humler, E., Langmuir, C. H., and Daux, V. (1999). Depth versus age: New perspectives from the chemical
compositions of ancient crust. Earth and Planetary Science Letters, 173(1-2):7–23.
Humphreys, E. R. and Niu, Y. (2009). On the composition of ocean island basalts (OIB): The effects of
lithospheric thickness variation and mantle metasomatism. Lithos, 112:118–136.
Hurley, P. M. and Rand, J. R. (1969). Pre-drift continental nuclei. Science, 164:1229–1242.
Ingle, S., Ito, G., Mahoney, J. J., Chazey III., W., Sinton, J., Rotella, M., and Christie, D. M. (2010).
Mechanisms of geochemical and geophysical variations along the western Galápagos Spreading Center.
Geochemistry Geophysics Geosystems, 11(4):Q04003.
Ito, G. and Mahoney, J. J. (2005a). Flow and melting of a heterogeneous mantle: 1. Method and importance to the geochemistry of ocean island and mid-ocean ridge basalts. Earth and Planetary Science
Letters, 230(1-2):29–46.
Ito, G. and Mahoney, J. J. (2005b). Flow and melting of a heterogeneous mantle: 2. Implications for a
chemically nonlayered mantle. Earth and Planetary Science Letters, 230(1-2):47–63.
Janney, P. E. and Castillo, P. R. (1997). Geochemistry of Mesozoic Pacific mid-ocean ridge basalt:
Constraints on melt generation and the evolution of the Pacific upper mantle. Journal of Geophysical
Research, 102(B3):5207–5229.
Janney, P. E. and Castillo, P. R. (2001). Geochemistry of the oldest Atlantic oceanic crust suggests
mantle plume involvement in the early history of the central Atlantic Ocean. Earth and Planetary
Science Letters, 192(3):291–302.
Janney, P. E., Macdougall, J. D., Natland, J. H., and Lynch, M. A. (2000). Geochemical evidence from
the Pukapuka volcanic ridge system for a shallow enriched mantle domain beneath the South Pacific
Dissertation P.A. Brandl
119
References
Superswell. Chemical Geology, 181(1-2):47–60.
Jarosewich, E., Nelen, J. A., and Norberg, J. A. (1980). Reference Samples for Electron Microprobe
Analysis. Geostandards Newsletter, 4(1):43–47.
Jaupart, C. and Mareschal, J. C. (2010). Heat generation and transport in the Earth. Cambridge University Press, Cambridge.
Jenner, F. E. and O’Neill, H. S. C. (2012). Analysis of 60 elements in 616 ocean floor basaltic glasses.
Geochemistry Geophysics Geosystems, 13:Q02005.
Jochum, K. P., Hofmann, A. W., Ito, E., Seufert, H. M., and White, W. M. (1983). K, U and Th in midocean ridge basalt glasses and heat production, K/U and K/Rb in the mantle. Nature, 306(5942):431–
436.
Johnson, K. T. M., Dick, H. J. B., and Shimizu, N. (1990). Melting in the oceanic upper mantle: An ion
microprobe study of diopsides in abyssal peridotites. Journal of Geophysical Research, 95(B3):2661–
2678.
Jull, M., Kelemen, P. B., and Sims, K. W. W. (2002). Consequences of diffuse and channelled porous
melt migration on uranium series disequilibria. Geochimica et Cosmochimica Acta, 66(23):4133–4148.
Karson, J. A. (2002). Geologic Structure of the uppermost oceanic crust created at fast- to intermediaterate spreading centers. Annual Review of Earth and Planetary Sciences, 30(1):347–384.
Kasting, J. F., Howard, M. T., Wallmann, K., Veizer, J., Shields, G., and Jaffrés, J. (2006). Paleoclimates,
ocean depth, and the oxygen isotopic composition of seawater. Chemical Geology, 252(1-2):82–93.
Kelemen, P. B. (1990). Reaction between ultramafic rock and fractionating basaltic magma I. Phase
relations, the origin of calc-alkaline magma series, and the formation of discordant dunite. Journal of
Petrology, 31(1):51–98.
Kelemen, P. B., Braun, M., and Hirth, G. (2000). Spatial distribution of melt conduits in the mantle
beneath oceanic spreading ridges: Observations from the Ingalls and Oman ophiolites. Geochemistry
Geophysics Geosystems, 1(7):1005.
Kelemen, P. B., Dick, H. J. B., and Quick, J. E. (1992). Formation of harzburgite by pervasive melt/rock
reaction in the upper mantle. Nature, (358):635–641.
Kelemen, P. B., Koga, K., and Shimizu, N. (1997). Geochemistry of gabbro sills in the crust-mantle
transition zone of the Oman ophiolite: Implications for the origin of the oceanic lower crust. Earth
and Planetary Science Letters, 146(3-4):475–488.
Kelley, K. A., Plank, T., Grove, T. L., Stolper, E. M., Newman, S., and Hauri, E. (2006). Mantle
melting as a function of water content beneath back-arc basins. Journal of Geophysical Research,
111(B09208):1–27.
Klein, E. M. and Langmuir, C. H. (1987). Gobal correlations of ocean ridge basalt chemistry with axial
depth and crustal thickness. Journal of Geophysical Research, 92(B8):8089–8115.
Klemme, S., Blundy, J. D., and Wood, B. J. (2002). Experimental constraints on major and trace element
partitioning during partial melting of eclogite. Geochimica et Cosmochimica Acta, 66(17):3109–3123.
Kogiso, T., Hirschmann, M. M., and Reiners, P. W. (2004). Length scales of mantle heterogeneities and
their relationship to ocean island basalt geochemistry. Geochimica et Cosmochimica Acta, 68(2):345–
360.
Konter, J. G., Staudigel, H., Blichert-Toft, J., Hanan, B. B., Polvé, M., Davies, G. R., Shimizu, N., and
Schiffman, P. (2009). Geochemical stages at Jasper Seamount and the origin of intraplate volcanoes.
Geochemistry Geophysics Geosystems, 10(2):Q02001.
Koppers, A. A. P. (2002). ArArCALC–software for 40 Ar/39 Ar age calculations. Computers & Geosciences,
28(5):605–619.
Koppers, A. A. P. (2003). High-resolution 40 Ar/39 Ar dating of the oldest oceanic basement basalts in
the western Pacific basin. Geochemistry Geophysics Geosystems, 4(11):8914.
Korenaga, J. (2008). Urey ratio and the structure and evolution of Earth’s mantle. Reviews of Geophysics,
46:2007RG000241.
120
Dissertation P.A. Brandl
References
Kushiro, I. (1972). Effect of water on the composition of magmas formed at high pressures. Journal of
Petrology, 13(2):311–334.
Labrosse, S. and Jaupart, C. (2007). Thermal evolution of the Earth: Secular changes and fluctuations
of plate characteristics. Earth and Planetary Science Letters, 260(3):465–481.
Langmuir, C. H., Klein, E. M., and Plank, T. (1992). Petrological systematics of Mid-ocean ridge basalts:
Constraints on melt generation beneath ocean ridges. In Phipps Morgan, J., Blackman, D. K., and
Sinton, J. M., editors, Mantle flow and melt generation at mid-ocean ridges, pages 183–280. American
Geophysical Union, Washington, DC.
Langmuir, C. H., Vocke Jr., R. D., Hanson, G. N., and Hart, S. R. (1978). A general mixing equation
with applications to Icelandic basalts. Earth and Planetary Science Letters, 37(3):380–392.
Larson, R. L. (1991a). Geological consequences of superplumes. Geology, 19(10):963–966.
Larson, R. L. (1991b). Latest pulse of Earth: Evidence for a mid-Cretaceous superplume. Geology,
19(6):547–550.
Le Maitre, R. W., Bateman, P., Dudek, A., Keller, J., Lameyre le Bas, M. J., Sabine, P. A., Schmid,
R., Sorensen, H., Streckeisen, A., Wolley, A. R., and Zanettin, B. (1989). A classification of igneous
rocks and glossary of terms. Recommendations of the International Union of Geological Sciences
Subcommission on the Systematics of Igneous Rocks. Recommendations of the International Union of
Geological Sciences Subcommission on the Systematics of Igneous Rocks, Oxford.
Lenardic, A., Moresi, L., Jellinek, A., and Manga, M. (2005). Continental insulation, mantle cooling, and
the surface area of oceans and continents. Earth and Planetary Science Letters, 234(3-4):317–333.
Ligi, M., Bonatti, E., Bortoluzzi, G., Cipriani, A., Cocchi, L., Tontini, F. C., Carminati, E., Ottolini, L.,
and Schettino, A. (2012). Birth of an ocean in the Red Sea: Initial pangs. Geochemistry Geophysics
Geosystems, 13:Q08009.
Lin, J., Purdy, G. M., Schouten, H., Sempere, J. C., and Zervas, C. (1990). Evidence from gravity data
for focused magmatic accretion along the Mid-Atlantic Ridge. Nature, 344(6267):627–632.
Liu, C. Z., Snow, J. E., Hellebrand, E., Brügmann, G., von der Handt, A., Büchl, A., and Hofmann,
A. W. (2008). Ancient, highly heterogeneous mantle beneath Gakkel ridge, Arctic Ocean. Nature,
452(7185):311–316.
Lundstrom, C. (2000). Models of U-series disequilibria generation in MORB: The effects of two scales of
melt porosity. Physics of The Earth and Planetary Interiors, 121(3-4):189–204.
Lundstrom, C. C., Gill, J., Williams, Q., and Hanan, B. B. (1998). Investigating solid mantle upwelling
beneath mid-ocean ridges using U-series disequilibria. II. A local study at 33 ◦ S Mid-Atlantic Ridge.
Earth and Planetary Science Letters, 157(3-4):167–181.
Macdonald, K. C. (1982). Mid-ocean ridges: Fine scale tectonic, volcanic and hydrothermal processes
within the plate boundary zone. Annual Review of Earth and Planetary Sciences, 10(1):155–190.
Macdonald, K. C., Fox, P. J., Perram, L. J., Eisen, M. F., Haymon, R. M., Miller, S. P., Carbotte, S. M.,
Cormier, M.-H., and Shor, A. N. (1988). A new view of the mid-ocean ridge from the behaviour of
ridge-axis discontinuities. Nature, 335:217–225.
Machetel, P. and Humler, E. (2003). High mantle temperature during Cretaceous avalanche. Earth and
Planetary Science Letters, 208(3-4):125–133.
Maclennan, J. (2008). Lead isotope variability in olivine-hosted melt inclusions from Iceland. Geochimica
et Cosmochimica Acta, 72(16):4159–4176.
Maclennan, J., McKenzie, D., Hilton, F., Grönvold, K., and Shimizu, N. (2003). Geochemical variability
in a single flow from northern Iceland. Journal of Geophysical Research, 108(B1):4–1–4–21.
Magde, L. S., Barclay, A. H., Toomey, D. R., Detrick, R. S., and Collins, J. A. (2000). Crustal magma
plumbing within a segment of the Mid-Atlantic Ridge, 35 ◦ N. Chemical Geology, 175(1-2):55–67.
Maicher, D., White, J. D. L., and Batiza, R. (2000). Sheet hyaloclastite: Density-current deposits of
quench and bubble-burst fragments from thin, glassy sheet lava flows, Seamount Six, Eastern Pacific
Ocean. Marine Geology, 171(1):75–94.
Dissertation P.A. Brandl
121
References
Mammerickx, J., Herron, E., and Dorman, L. (1980). Evidence for two fossil spreading ridges in the
southeast Pacific. Geological Society of America Bulletin, 91:263–271.
Manhes, G., Allègre, C. J., Dupré, B., and Hamelin, B. (1980). Lead isotope study of basic-ultrabasic
layered complexes: Speculations about the age of the earth and primitive mantle characteristics. Earth
and Planetary Science Letters, 47(3):370–382.
Marty, B. and Tolstikhin, I. N. (1998). CO 2 fluxes from mid-ocean ridges, arcs and plumes. Chemical
Geology, 145(3-4):233–248.
Marty, J. C. and Cazenave, A. (1989). Regional variations in subsidence rate of oceanic plates: A global
analysis. Chemical Geology, 94(3-4):301–315.
McCulloch, M. T. and Bennett, V. C. (1994). Progressive growth of the Earth’s continental crust and
depleted mantle: Geochemical constraints. Chemical Geology, 58(21):4717–4738.
McDade, P., Blundy, J. D., and Wood, B. J. (2003a). Trace element partitioning between mantle wedge
peridotite and hydrous MgO-rich melt. American Mineralogist, 88:1825–1831.
McDade, P., Blundy, J. D., and Wood, B. J. (2003b). Trace element partitioning on the Tinaquillo
Lherzolite solidus at 1.5 GPa. Physics of the Earth and Planetary Interiors, 139(1-2):129–147.
McDonough, W. F. and Sun, S. S. (1995). The composition of the Earth. Chemical Geology, 120(34):223–253.
McKenzie, D. and Bickle, M. J. (1988). The volume and composition of melt generated by extension of
the lithosphere. Journal of Petrology, 29(3):625–679.
McKenzie, D. and O’Nions, R. K. (1991). Partial melt distributions from inversion of rare earth element
concentrations. Journal of Petrology, 32(5):1021–1091.
McNutt, M. (1986). Nonuniform magnetization of seamounts: A least squares approach. Journal of
Geophysical Research, 91(B3):3686–3700.
McNutt, M. and Batiza, R. (1981). Paleomagnetism of northern Cocos seamounts: Constraints on absolute
plate motion. Geology, 9(4):148–154.
Meibom, A. and Anderson, D. L. (2004). The statistical upper mantle assemblage. Earth and Planetary
Science Letters, 217(1-2):123–139.
Menard, H. W., Chase, T. E., and Smith, S. M. (1964). Galapagos rise in the southeastern Pacific. Deep
Sea Research and Oceanographic Abstracts, 11(2):233–242.
Michael, P. J. and Cornell, W. C. (1998). Influence of spreading rate and magma supply on crystallization
and assimilation beneath mid-ocean ridges: Evidence from chlorine and major element chemistry of
mid-ocean ridge basalts. Journal of Geophysical Research, 103(B8):18325–18356.
Minshull, T. A., Bruguier, N. J., and Brozena, J. M. (1998). Ridge–plume interactions or mantle heterogeneity near Ascension Island? Geology, 26(2):115–118.
Miyashiro, A., Shido, F., and Ewing, M. (1969). Diversity and origin of abyssal tholeiite from the
Mid-Atlantic Ridge near 24 ◦ and 30 ◦ North latitude. Contributions to Mineralogy and Petrology,
23(1):38–52.
Moore, J. G. (1970). Water Content of Basalt Erupted on the ocean floor. Contributions to Mineralogy
and Petrology, 28(4):272–279.
Morgan, W. J. (1972). Deep mantle convection plumes and plate motions. AAPG Bulletin, 56(2):203–213.
Müller, R. D., Sdrolias, M., Gaina, C., and Roest, W. R. (2008a). Age, spreading rates, and spreading
asymmetry of the world’s ocean crust. Geochemistry Geophysics Geosystems, 9(4):Q04006.
Müller, R. D., Sdrolias, M., Gaina, C., Steinberger, B., and Heine, C. (2008b). Long-term sea-level
fluctuations driven by ocean basin dynamics. Science, 319(5868):1357–1362.
Mutter, C. and Mutter, J. (1993). Variations in thickness of layer 3 dominate oceanic crustal structure.
Earth and Planetary Science Letters, 117(1):295–317.
Nauret, F., Abouchami, W., Galer, S. J. G., Hofmann, A. W., Hémond, C., Chauvel, C., and Dyment,
J. (2006). Correlated trace element-Pb isotope enrichments in Indian MORB along 18–20 ◦ S, Central
122
Dissertation P.A. Brandl
References
Indian Ridge. Earth and Planetary Science Letters, 245(1-2):137–152.
Niu, Y. (2004). Bulk-rock major and trace element compositions of abyssal peridotites: Implications
for mantle melting, melt extraction and post-melting processes beneath mid-ocean ridges. Journal of
Petrology, 45(12):2423–2458.
Niu, Y. and Batiza, R. (1997). Trace element evidence from seamounts for recycled oceanic crust in the
Eastern Pacific mantle. Earth and Planetary Science Letters, 148(3-4):471–483.
Niu, Y., Bideau, D., Hékinian, R., and Batiza, R. (2001). Mantle compositional control on the extent of
mantle melting, crust production, gravity anomaly, ridge morphology, and ridge segmentation: A case
study at the Mid-Atlantic Ridge 33–35 ◦ N. Chemical Geology, 186(3-4):383–399.
Niu, Y., Collerson, K. D., Wendt, J. I., Regelous, M., and Batiza, R. (1999). Origin of enriched-type
mid-ocean ridge basalt at ridges far from mantle plumes – the East Pacific Rise at 11 ◦ 20’N. Journal
of Geophysical Research, 104(B4):7067–7087.
Niu, Y. and O’Hara, M. J. (2003). Origin of ocean island basalts: A new perspective from petrology,
geochemistry, and mineral physics considerations. Journal of Geophysical Research, 108(B4):2209.
Niu, Y. and O’Hara, M. J. (2008). Global correlations of ocean ridge basalt chemistry with axial depth:
A new perspective. Journal of Petrology, 49(4):633–664.
Niu, Y., Regelous, M., Wendt, I. J., Batiza, R., and O’Hara, M. J. (2002). Geochemistry of near-EPR
seamounts: Importance of source vs. process and the origin of enriched mantle component. Earth and
Planetary Science Letters, 199(3-4):327–345.
O’Hara, M. J. (1965). Primary magmas and the origin of basalts. Scottish Journal of Geology, 1(1):19–40.
O’Hara, M. J. (1968). Are Ocean Floor Basalts Primary Magma? Nature, 220(5168):683–686.
O’Neill, H. S. C. and Jenner, F. E. (2013). The global pattern of trace-element distributions in ocean
floor basalts. Nature, 491(7426):698–704.
O’Neill, H. S. C., Rubie, D. C., Canil, D., Geiger, C. A., Ross, C. A., Seifert, F., and Woodland, A. B.
(1993). Ferric iron in the upper mantle and in transition zone assemblages: Implications for relative
oxygen fugacities in the mantle. In Takahashi, E., Jeanloz, R., and Rubie, D., editors, Evolution of the
Earth and Planets, pages 73–88. American Geophysical Union, Washington, DC.
O’Reilly, S. Y. and Griffin, W. L. (1988). Mantle metasomatism beneath western Victoria, Australia: I.
Metasomatic processes in Cr-diopside lherzolites. Chemical Geology, 52(2):433–447.
Parmentier, E. M. and Phipps Morgan, J. (1990). Spreading rate dependence of three-dimensional structure in oceanic spreading centres. Nature, 348(6299):325–328.
Pearce, J. A. (2005). Mantle preconditioning by melt extraction during flow: Theory and petrogenetic
implications. Journal of Petrology, 46(5):973.
Perfit, M. R. and Chadwick, W. W. (1998). Magmatism at mid-ocean ridges: Constraints from volcanological and geochemical investigations. In Buck, W. R., Delaney, P. T., Karson, J. A., and Lagabrielle, Y.,
editors, Faulting and magmatism at mid-ocean ridges, volume 106 of Geophysical Monograph Series,
pages 59–115. American Geophysical Union, Washington, DC.
Perfit, M. R., Fornari, D. J., Ridley, W. I., Kirk, P. D., Casey, J., Kastens, K. A., Reynolds, J. R.,
Edwards, M., Desonie, D., and Shuster, R. (1996). Recent volcanism in the Siqueiros transform fault:
Picritic basalts and implications for MORB magma genesis. Earth and Planetary Science Letters,
141(1):91–108.
Pertermann, M. and Hirschmann, M. M. (2003). Partial melting experiments on a MORB-like pyroxenite
between 2 and 3 GPa: Constraints on the presence of pyroxenite in basalt source regions from solidus
location and melting rate. Journal of Geophysical Research, 108(B2):2125.
Pertermann, M., Hirschmann, M. M., Hametner, K., Günther, D., and Schmidt, M. W. (2004). Experimental determination of trace element partitioning between garnet and silica-rich liquid during anhydrous
partial melting of MORB-like eclogite. Geochemistry Geophysics Geosystems, 5(5):Q05A01.
Phillips, B. R. and Coltice, N. (2010). Temperature beneath continents as a function of continental cover
and convective wavelength. Journal of Geophysical Research, 115(B04408):1–13.
Dissertation P.A. Brandl
123
References
Phipps Morgan, J. (2001). Thermodynamics of pressure release melting of a veined plum pudding mantle.
Geochemistry Geophysics Geosystems, 2:2000GC000049.
Phipps Morgan, J. and Chen, Y. J. (1993). Dependence of ridge-axis morphology on magma supply and
spreading rate. Nature, 364(6439):706–708.
Phipps Morgan, J. and Morgan, W. J. (1999). Two-stage melting and the geochemical evolution of the
mantle: A recipe for mantle plum-pudding. Earth and Planetary Science Letters, 170(3):215–239.
Phipps Morgan, J., Parmentier, E. M., and Lin, J. (1987). Mechanisms for the origin of mid-ocean ridge
axial topography – Implications for the thermal and mechanical structure of accreting plate boundaries.
Journal of Geophysical Research, 92(B2):12823–12836.
Pilet, S., Hernandez, J., Sylvester, P., and Poujol, M. (2005). The metasomatic alternative for ocean
island basalt chemical heterogeneity. Earth and Planetary Science Letters, 236(1-2):148–166.
Presnall, D. C., Dixon, J. R., O’Donnell, T. H., and Dixons, S. A. (1979). Generation of mid-ocean ridge
tholeiites. Journal of Petrology, 20:3–35.
Prinzhofer, A., Lewin, E., and Allègre, C. J. (1989). Stochastic melting of the marble cake mantle:
Evidence from local study of the East Pacific Rise at 12 ◦ 50’N. Chemical Geology, 92(2):189–206.
Purdy, G. M., Kong, L. S. L., Christeson, G. L., and Solomon, S. C. (1992). Relationship between
spreading rate and the seismic structure of mid-ocean ridges. Nature, 355:815–817.
Putirka, K. D. (2005). Mantle potential temperatures at Hawaii, Iceland, and the mid-ocean ridge system,
as inferred from olivine phenocrysts: Evidence for thermally driven mantle plumes. Geochemistry
Geophysics Geosystems, 6(5):Q05L08.
Putirka, K. D. (2008a). Excess temperatures at ocean islands: Implications for mantle layering and
convection. Geology, 36(4):283–286.
Putirka, K. D. (2008b). Thermometers and Barometers for Volcanic Systems. In Putirka, K. D. and
Tepley, F., editors, Reviews in Mineralogy and Geochemistry, pages 61–120.
Putirka, K. D., Perfit, M. R., Ryerson, F. J., and Jackson, M. G. (2007). Ambient and excess mantle
temperatures, olivine thermometry, and active vs. passive upwelling. Chemical Geology, 241(3-4):177–
206.
Regelous, M., Hofmann, A. W., Abouchami, W., and Galer, S. J. G. (2003). Geochemistry of lavas from
the Emperor Seamounts, and the geochemical evolution of Hawaiian magmatism from 85 to 42 Ma.
Journal of Petrology, 44(1):113–140.
Regelous, M., Niu, Y., Wendt, J. I., Batiza, R., Greig, A., and Collerson, K. D. (1999). Variations in the
geochemistry of magmatism on the East Pacific Rise at 10 ◦ 30’N since 800 ka. Earth and Planetary
Science Letters, 168(1-2):45–63.
Rhodes, J., Dungan, M., Blanchard, D., and Long, P. (1979). Magma mixing at mid-ocean ridges:
Evidence from basalts drilled near 22 ◦ N on the Mid-Atlantic Ridge. Tectonophysics, 55(1-2):35–61.
Richter, F. M. (1984). Regionalized models for the thermal evolution of the Earth. Chemical Geology,
68(3):471–484.
Rickwood, P. C. (1989). Boundary lines within petrologic diagrams which use oxides of major and minor
elements. Lithos, 22(4):247–263.
Roeder, P. L. and Emslie, R. F. (1970). Olivine-liquid equilibrium. Contributions to Mineralogy and
Petrology, 29(4):275–289.
Rolf, T., Coltice, N., and Tackley, P. J. (2012). Linking continental drift, plate tectonics and the thermal
state of the Earth’s mantle. Earth and Planetary Science Letters, 351-352:134–146.
Rubin, K. H. and Macdougall, J. D. (1990). Dating of neovolcanic MORB using (226 Ra/230 Th) disequilibrium. Earth and Planetary Science Letters, 101(2-4):313–322.
Rubin, K. H. and Sinton, J. M. (2007). Inferences on mid-ocean ridge thermal and magmatic structure
from MORB compositions. Earth and Planetary Science Letters, 260(1-2):257–276.
Rubin, K. H., Sinton, J. M., Maclennan, J., and Hellebrand, E. (2009). Magmatic filtering of mantle
compositions at mid-ocean-ridge volcanoes. Nature Geoscience, 2(5):321–328.
124
Dissertation P.A. Brandl
References
Rubin, K. H., van der Zander, I., Smith, M. C., and Bergmanis, E. C. (2005). Minimum speed limit for
ocean ridge magmatism from 210 Pb–226 Ra–230 Th disequilibria. Nature, 437(7058):534–538.
Russo, C. J., Rubin, K. H., and Graham, D. W. (2009). Mantle melting and magma supply to the
Southeast Indian Ridge: The roles of lithology and melting conditions from U-series disequilibria.
Earth and Planetary Science Letters, 278:55–66.
Ryan, W. B. F., Carbotte, S. M., Coplan, J., O’Hara, S., Melkonian, A., Arko, R., Weissel, R. A., Ferrini,
V., Goodwillie, A., Nitsche, F., Bonczkowski, J., and Zemsky, R. (2009). Global multi-resolution
topography synthesis. Geochemistry Geophysics Geosystems, 10:Q03014.
Salters, V. J. M. (1996). The generation of mid-ocean ridge basalts from the Hf and Nd isotope perspective.
Earth and Planetary Science Letters, 141(1-4):109–123.
Salters, V. J. M. and Dick, H. J. B. (2002). Mineralogy of the mid-ocean-ridge basalt source from
neodymium isotopic composition of abyssal peridotites. Nature, 418(6893):68–72.
Salters, V. J. M. and Hart, S. R. (1989). The hafnium paradox and the role of garnet in the source of
mid-ocean-ridge basalts. Nature, 342(6248):420–422.
Salters, V. J. M. and Longhi, J. (1999). Trace element partitioning during the initial stages of melting
beneath mid-ocean ridges. Earth and Planetary Science Letters, 166(1-2):15–30.
Salters, V. J. M., Longhi, J. E., and Bizimis, M. (2002). Near mantle solidus trace element partitioning
at pressures up to 3.4 GPa. Geochemistry Geophysics Geosystems, 3(7):2001GC000148.
Salters, V. J. M. and Stracke, A. (2004). Composition of the depleted mantle. Geochemistry Geophysics
Geosystems, 5(5):Q05B07.
Saunders, A. D., Norry, M. J., and Tarney, J. (1988). Origin of MORB and chemically-depleted mantle
reservoirs: Trace element constraints. Journal of Petrology, Special Volume 1:415–445.
Scheirer, D. S. and Macdonald, K. C. (1993). Variation in cross-sectional area of the axial ridge along the
East Pacific Rise: Evidence for the magmatic budget of a fast spreading center. Journal of Geophysical
Research, 98(B5):7871–7885.
Scheirer, D. S. and Macdonald, K. C. (1995). Near-axis seamounts on the flanks of the East Pacific Rise,
8 ◦ N to 17 ◦ N. Journal of Geophysical Research, 100(B2):2239–2259.
Schilling, J. G., Zajac, M., Evans, R., Johnston, T., White, W. M., Devine, J. D., and Kingsley, R. (1983).
Petrologic and geochemical variations along the Mid-Atlantic Ridge from 29 ◦ N to 73 ◦ N. American
Journal of Science, 283(6):510–586.
Sclater, J. G., Jaupart, C., and Galson, D. (1980). The heat flow through oceanic and continental crust
and the heat loss of the Earth. Reviews of Geophysics, 18(1):269–311.
Scott, D. R. and Stevenson, D. J. (1989). A self-consistent model of melting, magma migration and
buoyancy-driven circulation beneath mid-ocean ridges. Journal of Geophysical Research, 94(B3):2973–
2988.
Seton, M., Gaina, C., Müller, R. D., and Heine, C. (2009). Mid-Cretaceous seafloor spreading pulse: Fact
or fiction? Geology, 37(8):687–690.
Shen, Y. and Forsyth, D. W. (1995). Geochemical constraints on initial and final depths of melting
beneath mid-ocean ridges. Journal of Geophysical Research, 100(B2):2211–2237.
Shido, F., Miyashiro, A., and Ewing, M. (1971). Crystallization of abyssal tholeiites. Contributions to
Mineralogy and Petrology, 31(4):251–266.
Shimizu, N. (1998). The geochemistry of olivine-hosted melt inclusions in a FAMOUS basalt ALV519-4-1.
Physics of the Earth and Planetary Interiors, 107(1-3):183–201.
Sigmarsson, O., Condomines, M., Grönvold, K., and Thordarson, T. (1991). Extreme magma homogeneity
in the 1783–84 Lakagigar eruption: Origin of a large volume of evolved basalt in Iceland. Geophysical
Research Letters, 18(12):2229.
Sims, K. W. W., Blichert-Toft, J., Fornari, D. J., Perfit, M. R., Goldstein, S. J., Johnson, P., DePaolo,
D. J., Hart, S. R., Murell, M. T., Michael, P. J., Layne, G. D., and Ball, L. A. (2003). Aberrant youth:
Chemical and isotopic constraints on the origin of off-axis lavas from the East Pacific Rise, 9–10 ◦ N.
Dissertation P.A. Brandl
125
References
Geochemistry Geophysics Geosystems, 4(10):8621.
Sims, K. W. W., DePaolo, D. J., Murrell, M. T., Baldridge, W. S., Goldstein, S., Clague, D., and Jull,
M. (1999). Porosity of the melting zone and variations in the solid mantle upwelling rate beneath
Hawaii: Inferences from 238 U-230 Th-226 Ra and 235 U-231 Pa disequilibria. Geochimica et Cosmochimica
Acta, 63(23-24):4119–4138.
Sims, K. W. W., Goldstein, S. J., Blichert-Toft, J., Perfit, M. R., Kelemen, P., Fornari, D. J., Michael,
P., Murrell, M. T., Hart, S. R., DePaolo, D. J., Layne, G., Ball, L., Jull, M., and Bender, J. (2002).
Chemical and isotopic constraints on the generation and transport of magma beneath the East Pacific
Rise. Geochimica et Cosmochimica Acta, 66(19):3481–3504.
Sinton, J., Bergmanis, E., Rubin, K., Batiza, R., Gregg, T. K. P., Grönvold, K., Macdonald, K. C.,
and White, S. M. (2002). Volcanic eruptions on mid-ocean ridges: New evidence from the superfast
spreading East Pacific Rise, 17–19 ◦ S. Journal of Geophysical Research, 107(B6):3–1–3–14.
Sinton, J. M. and Detrick, R. S. (1992). Mid-ocean ridge magma chambers. Journal of Geophysical
Research, 97(B1):197–216.
Sleep, N. H. (1984). Tapping of magmas from ubiquitous mantle heterogeneities: An alternative to mantle
plumes? Journal of Geophysical Research, 89(B12):10,029–10,041.
Sleep, N. H. and Windley, B. F. (1982). Archean plate tectonics: Constraints and inferences. The Journal
of Geology, 90(4):363–379.
Small, C. (1998). Global systematics of mid-ocean ridge morphology.
Smith, D. K. and Cann, J. R. (1993). Building the crust at the Mid-Atlantic Ridge. Nature, 365:707–715.
Snow, J. E., Hart, S. R., and Dick, H. J. B. (1994). Nd and Sr isotope evidence linking mid-ocean-ridge
basalts and abyssal peridotites. Nature, 371(6492):57–60.
Sobolev, A. V., Hofmann, A. W., Sobolev, S. V., and Nikogosian, I. K. (2005). An olivine-free mantle
source of Hawaiian shield basalts. Nature, 434(7033):590–597.
Sobolev, A. V. and Shimizu, N. (1993). Ultra-depleted primary melt included in an olivine from the
Mid-Atlantic Ridge. Nature, 363(6425):151–154.
Sotin, C. and Parmentier, E. M. (1989). Dynamical consequences of compositional and thermal density
stratification beneath spreading centers. Geophysical Research Letters, 16(8):835–838.
Spiegelman, M. and Elliott, T. (1993). Consequences of melt transport for uranium series disequilibrium
in young lavas. Earth and Planetary Science Letters, 118(1-4):1–20.
Spiegelman, M. and Kelemen, P. B. (2003). Extreme chemical variability as a consequence of channelized
melt transport. Geochemistry Geophysics Geosystems, 4(7):1055.
Spiegelman, M. and McKenzie, D. (1987). Simple 2-D models for melt extraction at mid-ocean ridges
and island arcs. Earth and Planetary Science Letters, 83(1-4):137–152.
Stakes, D. S., Shervais, J. W., and Hopson, C. A. (1984). The volcanic-tectonic cycle of the FAMOUS
and AMAR Valleys, Mid-Atlantic Ridge (36 ◦ 47’N): Evidence from basalt glass and phenocryst compositional variations for a steady state magma chamber beneath the Valley Midsections, AMAR-3.
Journal of Geophysical Research, 89:6995–7028.
Standish, J. J. and Sims, K. W. W. (2010). Young off-axis volcanism along the ultraslow-spreading
Southwest Indian Ridge. Nature Geoscience, 3:286–292.
Staudigel, H., Davies, G. R., Hart, S. R., Marchant, K. M., and Smith, B. (1995). Large scale isotopic Sr,
Nd and O isotopic anatomy of altered oceanic crust: DSDP/ODP sites 417/418. Earth and Planetary
Science Letters, 130(1-4):169–185.
Staudigel, H., Muehlenbachs, K., Richardson, S. H., and Hart, S. R. (1981). Agents of low temperature
ocean crust alteration. Contributions to Mineralogy and Petrology, 77:150–157.
Stein, C. A. and Stein, S. (1992). A model for the global variation in oceanic depth and heat flow with
lithospheric age. Nature, 359(6391):123–129.
Stolper, E. and Asimow, P. (2007). Insights into mantle melting from graphical analysis of one-component
systems. American Journal of Science, 307(8):1051–1139.
126
Dissertation P.A. Brandl
References
Stracke, A. (2008). Chemical geodynamics: Tracking mantle depletion. Nature Geoscience, 1(4):215–216.
Stracke, A., Bizimis, M., and Salters, V. J. M. (2003). Recycling oceanic crust: Quantitative constraints.
Geochemistry Geophysics Geosystems, 4(3):8003.
Stracke, A. and Bourdon, B. (2009). The importance of melt extraction for tracing mantle heterogeneity.
Geochimica et Cosmochimica Acta, 73(1):218–238.
Stracke, A., Bourdon, B., and McKenzie, D. (2006). Melt extraction in the Earth’s mantle: Constraints
from U-Th-Pa-Ra studies in oceanic basalts. Earth and Planetary Science Letters, 244(1-2):97–112.
Stracke, A., Hofmann, A., and Hart, S. R. (2005). FOZO, HIMU, and the rest of the mantle zoo.
Geochemistry Geophysics Geosystems, 6(5):Q05007.
Stracke, A., Salters, V. J. M., and Sims, K. W. W. (1999). Assessing the presence of garnet-pyroxenite
in the mantle sources of basalts through combined hafnium-neodymium-thorium isotope systematics.
Geochemistry Geophysics Geosystems, 1(1):1006.
Stracke, A., Snow, J. E., Hellebrand, E., von der Handt, A., Bourdon, B., Birbaum, K., and Günther, D.
(2011). Abyssal peridotite Hf isotopes identify extreme mantle depletion. Earth and Planetary Science
Letters, 308(3-4):359–368.
Sturm, M. E., Goldstein, S. J., Klein, E. M., Karson, J. A., and Murrell, M. T. (2000). Uranium-series age
constraints on lavas from the axial valley of the Mid-Atlantic Ridge, MARK area. Chemical Geology,
181(1-2):61–70.
Su, W., Mutter, C. Z., Mutter, J. C., and Buck, W. R. (1994). Some theoretical predictions on the
relationships among spreading rate, mantle temperature, and crustal thickness. Journal of Geophysical
Research, 99:3215–3215.
Suhr, G. (1999). Melt migration under oceanic ridges: Inferences from reactive transport modelling of
upper mantle hosted dunites. Journal of Petrology, 40(4):575–599.
Suhr, G., Hellebrand, E., Snow, J. E., Seck, H. A., and Hofmann, A. W. (2003). Significance of large,
refractory dunite bodies in the upper mantle of the Bay of Islands Ophiolite. Geochemistry Geophysics
Geosystems, 4(3):8605.
Tatsumoto, M. (1978). Isotopic composition of lead in oceanic basalt and its implication to mantle
evolution. Earth and Planetary Science Letters, 38(1):63–87.
Taylor, B. and Martinez, F. (2003). Back-arc basin basalt systematics. Earth and Planetary Science
Letters, 210(3):481–497.
Taylor, S. R. and McLennan, S. M. (1995). The geochemical evolution of the continental crust. Reviews
of Geophysics, 33(2):241–465.
Taylor Jr., H. P. (1980). The effects of assimilation of country rocks by magmas on 18 O/16 O and 87 Sr/86 Sr
systematics in igneous rocks. Earth and Planetary Science Letters, 47(2):243–254.
The MELT Seismic Team (1998). Imaging the deep seismic structure beneath a mid-ocean ridge: The
MELT experiment. Science, 280:1215–1218.
Thompson, R. N. and Gibson, S. A. (2000). Transient high temperatures in mantle plume heads inferred
from magnesian olivines in Phanerozoic picrites. Nature, 407(6803):502–506.
Thordarson, T., Self, S., Óskarsson, N., and Hulsebosch, T. (1996). Sulfur, chlorine, and fluorine degassing
and atmospheric loading by the 1783-1784 AD Laki (Skaftár Fires) eruption in Iceland. Bulletin of
Volcanology, 58(2-3):205–225.
Tian, L., Castillo, P. R., Lonsdale, P. F., Hahm, D., and Hilton, D. R. (2011). Petrology and Sr-Nd-Pb-He
isotope geochemistry of postspreading lavas on fossil spreading axes off Baja California Sur, Mexico.
Geochemistry Geophysics Geosystems, 12:Q0AC10.
Todt, W., Cliff, R. A., Hanser, A., and Hofmann, A. W. (1996). Evaluation of a 202 Pb-205 Pb double spike
for high-precision lead isotope analysis . In Hart, S. and Basu, A., editors, Earth Processes: Reading
the isotopic code, pages 429–437. American Geophysical Union, Washington, DC.
Tormey, D. R., Grove, T. L., and Bryan, W. B. (1987). Experimental petrology of normal MORB near
the Kane Fracture Zone: 22–25 ◦ N, Mid-Atlantic Ridge. Contrib Mineral Petrol, 96:121–139.
Dissertation P.A. Brandl
127
References
Tryggvason, K., Husebye, E. S., and Stefánsson, R. (1983). Seismic image of the hypothesized Icelandic
hot spot. Tectonophysics, 100(1):97–118.
Tucholke, B. E., Lin, J., Kleinrock, M. C., Tivey, M. A., Reed, T. B., Goff, J., and Jaroslow, G. E. (1997).
Segmentation and crustal structure of the western Mid-Atlantic Ridge flank, 25 ◦ 25’–27 ◦ 10’N and 0–29
m.y. Journal of Geophysical Research, 102(B5):10203–10223.
Turner, S. P., Beier, C., Niu, Y., and Cook, C. (2011). U-Th-Ra disequilibria and the extent of off-axis
volcanism across the East Pacific Rise at 9 ◦ 30’N, 10 ◦ 30’N, and 11 ◦ 20’N. Geochemistry Geophysics
Geosystems, 12(7):Q0AC12.
Urey, H. C. (1956). The cosmic abundances of potassium, uranium, and thorium and the heat balances
of the Earth, the Moon, and Mars. Proceedings of the National Academy of Sciences, 42(12):889–891.
Walker, D., Shibata, T., and DeLong, S. E. (1979). Abyssal tholeiites from the Oceanographer Fracture
Zone. Contributions to Mineralogy and Petrology, 70(2):111–125.
Warren, J. M., Shimizu, N., Sakaguchi, C., Dick, H. J. B., and Nakamura, E. (2009). An assessment of
upper mantle heterogeneity based on abyssal peridotite isotopic compositions. Journal of Geophysical
Research, 114:B12203.
Wasserburg, G. J., MacDonald, G. J. F., Hoyle, F., and Fowler, W. A. (1964). Relative contributions of
uranium, thorium, and potassium to heat production in the Earth. Science, 143(3605):465–467.
Waters, C. L., Sims, K. W. W., Perfit, M. R., Blichert-Toft, J., and Blusztajn, J. (2011). Perspective on
the genesis of E-MORB from chemical and isotopic heterogeneity at 9–10 ◦ N East Pacific Rise. Journal
of Petrology, 52(3):565–602.
Wendt, J. I., Regelous, M., Niu, Y., Hékinian, R., and Collerson, K. D. (1999). Geochemistry of lavas from
the Garrett Transform Fault: Insights into mantle heterogeneity beneath the eastern Pacific. Earth
and Planetary Science Letters, 173(3):271–284.
Wessel, P. and Smith, W. H. F. (1991). Free software helps map and display data. Eos Transactions
AGU, 72(41):441 & 445–446.
Wessel, P. and Smith, W. H. F. (1995). New version of the generic mapping tools released. Eos Transactions AGU, 76(33):329.
Wessel, P. and Smith, W. H. F. (1998). New, improved version of generic mapping tools released. Eos
Transactions AGU, 79(47):579.
White, R. S. and McKenzie, D. (1989). Magmatism at rift zones: The generation of volcanic continental
margins and flood basalts. Journal of Geophysical Research, 94(B6):7685–7729.
White, R. S. and McKenzie, D. (1995). Mantle plumes and flood basalts. Journal of Geophysical Research,
100(B9):17543–17585.
White, R. S., McKenzie, D., and O’Nions, R. K. (1992). Oceanic crustal thickness from seismic measurements and rare earth element inversions. Journal of Geophysical Research, 97(B13):19683–19715.
White, W. M. and Hofmann, A. W. (1982). Sr and Nd isotope geochemistry of oceanic basalts and mantle
evolution. Nature, 296(5860):821–825.
White, W. M. and Schilling, J. G. (1978). The nature and origin of geochemical variation in Mid-Atlantic
Ridge basalts from the central North Atlantic. Geochimica et Cosmochimica Acta, 42:1501–1516.
Whittaker, J. M., Müller, R. D., Roest, W. R., Wessel, P., and Smith, W. H. F. (2008). How supercontinents and superoceans affect seafloor roughness. Nature, 456(7224):938–941.
Williamson, J. H. (1968).
46(16):1845–1847.
Least-squares fitting of a straight line.
Canadian Journal of Physics,
Winter, J. D. (2001). An introduction to igneous and metamorphic petrology. Prentice-Hall, New Yersey.
Wolfe, C. J., Bjarnason, I. T., VanDecar, J. C., and Solomon, S. C. (1997). Seismic structure of the
Iceland mantle plume. Nature, 385(6613):245–247.
Workman, R. K. and Hart, S. R. (2005). Major and trace element composition of the depleted MORB
mantle (DMM). Earth and Planetary Science Letters, 231(1-2):53–72.
128
Dissertation P.A. Brandl
References
Wyllie, P. J. (1971). Role of water in magma generation and initiation of diapiric uprise in the mantle.
Journal of Geophysical Research, 76(5):1328.
Wyllie, P. J. (1973). Experimental petrology and global tectonics – a preview. Chemical Geology,
17(3):189–209.
Wyllie, P. J. (1977). Peridotite–CO2 –H2 O, and carbonatitic liquids in the upper asthenosphere. Nature,
266:45–47.
Yaxley, G. M. and Brey, G. P. (2004). Phase relations of carbonate-bearing eclogite assemblages from 2.5
to 5.5 GPa: Implications for petrogenesis of carbonatites. Contributions to Mineralogy and Petrology,
146(5):606–619.
Yeo, I., Searle, R. C., Achenbach, K. L., Le Bas, T. P., and Murton, B. J. (2012). Eruptive hummocks:
Building blocks of the upper ocean crust. Geology, 40(1):91–94.
Zhong, S. and Gurnis, M. (1993). Dynamic feedback between a continentlike raft and thermal convection.
Journal of Geophysical Research, 98(B7):12219–12232.
Zindler, A. and Hart, S. (1986). Chemical geodynamics. Annual Review of Earth and Planetary Sciences,
14(1):493–571.
Zindler, A., Staudigel, H., and Batiza, R. (1984). Isotope and trace element geochemistry of young
Pacific seamounts: Implications for the scale of upper mantle heterogeneity. Earth and Planetary
Science Letters, 70(2):175–195.
Dissertation P.A. Brandl
129
Appendix
Chapter 5: High mantle temperatures following rifting
Table
Table
Table
Table
Table
Table
A1:
A2:
A3:
A4:
A5:
A6:
Overview of sampled ocean drilling sites.
Major element composition of ancient Atlantic MORB.
Major element composition of ancient Pacific MORB.
Results of electron microprobe standard analyses.
Averages for 500 m ridge depth intervals of the global mid-ocean ridge system.
Fractionation corrected data of ancient MORB with MgO ≥7.0 wt. % normalised
to MgO = 8.0 wt. %.
Chapter 6: Volcanism on the flanks of the East Pacific Rise
Table A7:
Major element, trace element and Sr-Nd-Pb isotope analyses of Seamount 6
lavas.
Chapter 7: Post-spreading volcanism on the fossil Galapagos Rise
Table A8:
Geochemical and isotopic composition of the Galapagos Rise lavas.
Chapter 8: Compositional variation of lavas from a young volcanic field
Table A9:
Major and trace element and radiogenic isotope data of lavas from the 8◦ 48’S
volcanic field.
Table A1: Overview of sampled ocean drilling sites.
Site
Ocean
WD
Basement
[Ma]
Age
Latitude Longitude
[°N]
[°E]
[m]
[m]
n
comments
DSDP leg 11, site 105
Atlantic
156
34.90
-69.17
5251
5875
8
DSDP leg 37, site 335*
Atlantic
16.5
37.30
-35.20
3188
3642
53
DSDP leg 45, site 396A
Atlantic
13
22.98
-43.52
4450
4575
15
DSDP leg 49, site 410A*
Atlantic
10.7
45.51
-29.48
2977
3310
8
DSDP leg 51-53, site 417D
Atlantic
120
25.11
-68.05
5842
5825
55
DSDP 51-53, leg 418A
Atlantic
120
25.04
-68.06
5514
5838
13
DSDP leg 73, site 522B
Atlantic
37.1
-26.11
-5.13
4441
4595
12
DSDP leg 76, site 534A
Atlantic
162
28.34
-75.38
4971
6606
4
DSDP leg 78, site 543A
Atlantic
80
15.71
-58.65
5633
6044
15
DSDP leg 16, site 163
Pacific
72
11.24
-150.29
5230
5507
2
DSDP leg 17, site 166
Pacific
115
3.76
-175.08
4962
5269
2
DSDP leg 29, site 278
Pacific
30
-56.56
160.07
3708
4137
5
DSDP leg 63, site 469
Pacific
17
32.62
-120.55
3790
4188
7
DSDP leg 63, site 470A
Pacific
15.5
29.09
-117.52
3549
3716
13
DSDP leg 63, site 472
Pacific
15
23.01
-114.00
3831
3943
1
DSDP leg 68, site 501
Pacific
5.9
1.23
-83.73
3457
3721
8
Costa Rica Rift
DSDP leg 69, site 504B
Pacific
5.9
1.23
-83.73
3460
3738
22
Costa Rica Rift
DSDP leg 85, site 573B
Pacific
35.5
0.50
-133.31
4301
4829
1
DSDP leg 91, site 595B
Pacific
90
-23.82
-165.53
5616
5697
2
DSDP leg 92, site 597A
Pacific
28.3
-18.81
-129.77
4163
4211
1
ODP leg 185, site 801C
Pacific
166
18.65
156.37
5685
6180
9
ODP leg 185, site 1149D
Pacific
132
31.31
143.40
5929
6149
1
ODP leg 191, site 1179D
Pacific
129
41.08
159.96
5575
5952
11
ODP leg 199, site 1215B
Pacific
58
26.03
-147.93
5398
5472
1
ODP leg 199, site 1217A
Pacific
48
16.87
-138.10
5342
5480
2
ODP leg 199, site 1222A
Pacific
56
13.82
-143.89
4989
5087
2
ODP leg 200, site 1224F
Pacific
46
27.89
-141.98
4967
4995
8
ODP leg 203, site 1243B
Pacific
10
5.30
-110.07
3882
3991
2
ODP leg 206, site 1256C
Pacific
15
6.74
-91.93
3635
3888
6
"Superfast"
ODP leg 206, site 1256D
Pacific
15
6.74
-91.93
3635
3911
51
"Superfast"
S' of Azores platform
MAR 45°N melting anomaly
E' of Southern MAR
Macquarie Ridge
* These sites are clearly influenced by melting anomalies. Thus they have been excluded from any calculation within this paper
but data are shown for completeness.
WD
Basement
n
Water depth
Depth to the top of oceanic igneous basement
Number of samples
Table A2: Major element composition of ancient Atlantic MORB (uncorrected raw data).
sample ID
DiB
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[m]
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP011-0105-041-002/117-119
0.67
50.08
1.02
15.27
9.64
0.18
8.48
13.09
1.87
0.07
0.09
1104
34
100.07
DSDP011-0105-041-003/46-48
2.08
49.90
1.01
15.10
9.61
0.19
9.18
12.75
1.86
0.07
0.08
1109
46
100.03
DSDP011-0105-041-003/108-110
1.46
50.12
0.99
15.07
9.68
0.19
9.03
12.81
1.88
0.07
0.09
1075
41
100.21
DSDP011-0105-042-001/50-52
2.22
50.45
0.99
15.06
9.56
0.17
9.06
12.75
1.89
0.07
0.08
1101
37
100.36
DSDP011-0105-042-001/122-124
1.50
50.61
1.01
15.02
9.66
0.17
9.13
12.86
1.86
0.07
0.09
1096
41
100.76
DSDP011-0105-042-002//88-89
3.38
50.25
1.02
15.18
9.73
0.19
9.04
12.83
1.85
0.07
0.07
1138
44
100.52
DSDP011-0105-043-001/77-79
5.36
49.87
1.01
15.04
9.64
0.18
9.09
12.83
1.88
0.07
0.08
1085
17
99.97
DSDP011-0105-043-002/25-26
5.75
50.27
1.00
15.09
9.62
0.18
9.13
12.82
1.89
0.07
0.06
1154
60
100.43
DSDP037-0335-005-002/65.5-66.5
0.15
50.25
1.17
15.74
9.27
0.19
8.42
13.02
2.39
0.18
0.11
1084
253
101.03
DSDP037-0335-005-003/37.5-39
1.38
48.84
1.17
15.63
9.34
0.20
8.61
12.85
2.40
0.17
0.11
1137
233
99.61
DSDP037-0335-006-001/46-48
7.96
49.42
1.16
15.65
9.36
0.19
8.66
12.79
2.42
0.17
0.11
1105
210
100.22
DSDP037-0335-006-002/90-93
9.90
49.24
1.17
15.59
9.32
0.18
8.60
12.70
2.40
0.17
0.12
1121
212
99.79
DSDP037-0335-006-003/58-61
11.08
49.85
1.15
15.58
9.32
0.18
8.58
12.95
2.41
0.18
0.13
1092
226
100.62
DSDP037-0335-006-004/71-75
12.71
49.93
1.17
15.62
9.34
0.18
8.56
12.96
2.39
0.17
0.11
1106
204
100.72
DSDP037-0335-006-005/51.5-53
14.02
49.86
1.17
15.75
9.31
0.18
8.55
12.91
2.41
0.17
0.09
1081
213
100.69
DSDP037-0335-006-006/29-31
15.29
49.88
1.16
15.49
9.33
0.19
8.47
12.81
2.39
0.17
0.11
1072
204
100.29
DSDP037-0335-006-CC/78-80
17.28
50.28
1.14
15.67
9.30
0.18
8.60
13.06
2.29
0.17
0.10
1113
253
101.10
DSDP037-0335-007-001/52-55
17.52
50.11
1.15
15.68
9.34
0.18
8.57
13.00
2.40
0.17
0.11
1121
207
101.02
DSDP037-0335-007-002/64-67
19.60
49.83
1.17
15.75
9.34
0.19
8.52
12.97
2.40
0.17
0.11
1089
228
100.73
DSDP037-0335-007-002/110-112
19.14
50.02
1.14
15.69
9.35
0.18
8.47
13.04
2.40
0.17
0.12
1074
228
100.89
DSDP037-0335-007-003/34-36
20.34
50.36
1.16
15.61
9.35
0.19
8.49
13.06
2.39
0.17
0.09
1025
232
101.15
DSDP037-0335-007-003/92-93
20.92
49.92
1.16
15.75
9.34
0.17
8.50
13.01
2.36
0.17
0.10
1035
218
100.77
DSDP037-0335-008-001/35-36
27.51
50.49
1.17
15.62
9.35
0.18
8.50
13.08
2.41
0.17
0.09
1023
236
101.35
DSDP037-0335-008-001/101-102
26.85
50.11
1.16
15.70
9.33
0.18
8.45
12.94
2.37
0.17
0.12
1030
212
100.81
DSDP037-0335-008-002/20-21
28.20
50.24
1.15
15.62
9.31
0.20
8.48
12.92
2.39
0.17
0.10
1034
255
100.85
DSDP037-0335-008-003/38-40
29.88
50.15
1.16
15.70
9.32
0.19
8.48
12.96
2.38
0.18
0.11
1060
253
100.91
DSDP037-0335-008-003/90-92
30.40
50.17
1.18
15.56
9.30
0.19
8.52
13.01
2.38
0.18
0.10
1106
257
100.89
DSDP037-0335-008-004/1-3
31.01
49.88
1.16
15.57
9.29
0.18
8.45
12.99
2.43
0.17
0.12
1053
213
100.52
DSDP037-0335-008-004/81-82
31.81
50.10
1.16
15.48
9.29
0.18
8.42
13.00
2.40
0.16
0.11
1032
246
100.59
DSDP037-0335-008-CC/18-20
32.68
50.45
1.16
15.51
9.34
0.18
8.48
13.02
2.42
0.17
0.11
1119
244
101.14
DSDP037-0335-008-CC/94-96
33.44
50.29
1.15
15.55
9.35
0.18
8.57
13.02
2.45
0.17
0.11
1023
184
101.12
DSDP037-0335-009-001/60-63
37.43
50.56
1.15
15.47
9.27
0.18
8.38
13.02
2.45
0.17
0.11
1133
207
101.07
DSDP037-0335-009-001/110-112
36.60
50.28
1.16
15.60
9.33
0.18
8.40
13.04
2.39
0.17
0.09
1083
235
100.93
DSDP037-0335-009-001/143-146
37.10
50.14
1.15
15.46
9.34
0.18
8.34
13.03
2.41
0.17
0.10
1101
199
100.61
DSDP037-0335-009-002/54-56
38.04
50.39
1.15
15.50
9.26
0.18
8.38
12.98
2.40
0.17
0.12
1046
193
100.82
DSDP037-0335-009-003/35-38
40.30
50.18
1.12
15.60
9.22
0.17
8.44
13.11
2.42
0.17
0.10
1088
218
100.83
DSDP037-0335-009-004/7-11
41.86
50.28
1.13
15.59
9.25
0.18
8.40
13.17
2.36
0.17
0.10
1097
229
100.93
DSDP037-0335-009-004/136-138
40.57
50.38
1.11
15.60
9.27
0.19
8.44
13.14
2.38
0.16
0.11
1030
189
101.06
DSDP037-0335-009-005/48-50
42.48
50.42
1.13
15.62
9.23
0.18
8.53
13.20
2.39
0.18
0.09
1087
244
101.27
DSDP037-0335-009-005/92-94
42.92
50.12
1.13
15.43
9.18
0.17
8.36
13.10
2.39
0.16
0.11
1136
230
100.46
DSDP037-0335-010-001/70-74
46.57
50.40
1.13
15.63
9.21
0.17
8.48
13.16
2.41
0.17
0.09
1069
249
101.14
DSDP037-0335-010-001/107-109
46.20
50.41
1.12
15.59
9.18
0.18
8.48
13.11
2.39
0.17
0.10
1043
199
101.01
DSDP037-0335-010-002/90-93
47.90
50.44
1.14
15.63
9.20
0.18
8.51
13.12
2.39
0.17
0.09
1063
213
101.15
Table A2: Major element composition of ancient Atlantic MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP037-0335-010-003/44-47
49.73
50.05
1.12
15.46
9.17
0.19
8.55
13.17
2.39
0.17
0.09
1060
230
100.64
DSDP037-0335-010-003/123-126
48.94
49.96
1.12
15.48
9.21
0.17
8.47
13.14
2.36
0.18
0.10
1061
240
100.47
DSDP037-0335-010-004/42-44
51.03
50.14
1.13
15.59
9.24
0.18
8.46
13.18
2.35
0.17
0.09
1087
193
100.83
DSDP037-0335-010-004/103-110
50.42
49.53
1.11
15.54
9.14
0.18
8.46
13.21
2.34
0.17
0.10
1111
246
100.08
DSDP037-0335-010-005/9-11
51.59
50.06
1.11
15.52
9.20
0.18
8.42
13.21
2.40
0.17
0.09
1051
248
100.66
DSDP037-0335-011-001/32-34
55.32
49.72
1.12
15.64
9.16
0.18
8.30
13.21
2.37
0.16
0.10
1104
220
100.25
DSDP037-0335-011-002/92-94
57.42
49.88
1.11
15.65
9.18
0.17
8.44
13.15
2.31
0.17
0.09
1078
235
100.44
DSDP037-0335-012-001/102-104
65.52
50.42
1.12
15.51
9.18
0.17
8.47
13.24
2.40
0.17
0.10
1103
213
101.08
DSDP037-0335-012-002/40-42
66.40
50.23
1.12
15.63
9.19
0.18
8.38
13.29
2.37
0.17
0.09
1045
210
100.94
DSDP037-0335-013-001/15-17
74.44
50.04
1.14
15.64
9.22
0.17
8.29
13.20
2.36
0.17
0.10
1070
245
100.62
DSDP037-0335-013-001/44-46
75.29
50.23
1.13
15.58
9.20
0.18
8.43
13.22
2.41
0.17
0.11
1036
224
100.94
DSDP037-0335-013-001/129-131
74.15
50.07
1.13
15.66
9.18
0.18
8.31
13.21
2.36
0.17
0.10
1016
206
100.64
DSDP037-0335-013-002/35-37
76.65
50.15
1.13
15.65
9.23
0.18
8.25
13.18
2.39
0.17
0.10
1033
231
100.72
DSDP037-0335-013-002/115-118
75.85
50.16
1.14
15.68
9.22
0.17
8.36
13.18
2.38
0.17
0.11
1102
173
100.87
DSDP037-0335-013-003/54-56
77.54
50.28
1.13
15.52
9.29
0.19
8.23
13.21
2.37
0.18
0.11
996
224
100.79
DSDP037-0335-014-001/69-71
84.19
50.16
1.13
15.60
9.26
0.18
8.29
13.29
2.41
0.17
0.10
1029
223
100.88
DSDP037-0335-014-002/39-40
85.39
50.22
1.13
15.67
9.27
0.19
8.25
13.19
2.38
0.18
0.10
1041
206
100.85
DSDP037-0335-014-003/62-64
87.12
50.34
1.14
15.71
9.23
0.17
8.34
13.18
2.36
0.17
0.10
1070
219
101.02
DSDP045-0396-014-006/14-15
30.34
49.50
1.65
15.26
9.51
0.18
7.43
11.09
2.58
0.13
0.15
1232
27
97.79
DSDP045-0396-015-001/44-46
32.06
50.27
1.36
15.82
8.78
0.18
8.26
11.44
2.61
0.10
0.12
1159
37
99.24
DSDP045-0396-015-001/136-137
31.14
50.39
1.65
15.15
9.63
0.19
7.50
11.16
2.67
0.12
0.17
1281
27
98.97
DSDP045-0396-015-003/28-30
33.98
50.08
1.36
15.85
8.73
0.17
8.29
11.50
2.56
0.11
0.12
1066
40
99.03
DSDP045-0396-015-004/100-102
36.20
50.99
1.35
16.06
8.75
0.17
8.22
11.54
2.52
0.12
0.14
1118
52
100.13
DSDP045-0396-016-002/80-83
42.40
50.66
1.36
15.97
8.65
0.15
8.29
11.45
2.62
0.11
0.12
1143
41
99.68
DSDP045-0396-016-004/56-57
45.16
50.32
1.36
15.96
8.71
0.17
8.31
11.43
2.63
0.10
0.12
1111
33
99.40
DSDP045-0396-018-001/38-40
58.98
49.65
1.36
15.96
8.71
0.16
8.17
11.60
2.51
0.11
0.14
1094
9
98.64
DSDP045-0396-018-CC/25-27
63.35
50.43
1.37
15.94
8.73
0.17
8.34
11.38
2.63
0.11
0.13
1047
37
99.51
DSDP045-0396-019-002/61-63
69.71
50.97
1.39
16.28
8.77
0.17
8.39
11.54
2.25
0.11
0.13
1092
32
100.29
DSDP045-0396-021-001/93-95
87.03
49.24
1.37
15.97
8.58
0.16
8.06
11.47
2.46
0.10
0.11
1129
56
97.81
DSDP045-0396-022-001/115-117
96.45
50.37
1.38
15.94
8.74
0.17
8.29
11.49
2.64
0.11
0.13
1016
41
99.52
DSDP045-0396-022-003/5-6
98.35
50.79
1.35
15.94
8.74
0.18
8.33
11.49
2.58
0.11
0.13
1054
32
99.92
DSDP045-0396-022-004/98-100
100.78
50.52
1.51
14.99
9.71
0.19
7.64
11.46
2.51
0.10
0.13
1218
45
99.07
DSDP045-0396-024-003/72-76
117.92
50.65
1.53
15.08
9.63
0.20
7.74
11.48
2.53
0.10
0.13
1304
49
99.39
DSDP049-0410A-001-007/5-7
2.70
51.35
1.38
15.74
8.18
0.17
7.16
11.17
2.77
0.61
0.19
1004
573
99.03
DSDP049-0410A-001-007/21-23
2.54
50.71
1.37
15.69
8.13
0.16
7.14
11.18
2.75
0.60
0.23
935
603
98.26
DSDP049-0410A-002-002/11-13
5.11
51.65
1.39
15.68
8.17
0.16
7.24
11.26
2.76
0.60
0.22
1020
596
99.44
DSDP049-0410A-002-002/120-130
6.20
50.29
1.39
15.70
8.13
0.16
7.14
11.21
2.73
0.61
0.21
1003
589
97.87
DSDP049-0410A-002-003/50-51
7.00
51.89
1.40
15.77
8.19
0.16
7.13
11.28
2.67
0.61
0.21
1019
597
99.60
DSDP049-0410A-002-004/124-126
9.24
52.12
1.42
15.96
8.22
0.15
7.18
11.43
2.55
0.63
0.21
1052
549
100.18
DSDP049-0410A-003-001/135-136
13.49
50.24
1.39
15.86
8.12
0.17
7.13
11.17
2.72
0.60
0.23
996
577
97.93
DSDP049-0410A-004-002/27-29
24.27
49.34
1.42
16.45
7.89
0.18
7.41
11.60
2.68
0.78
0.27
995
816
98.35
DSDP051-0417D-022-001/95-96
136.95
50.00
1.61
14.35
11.20
0.23
7.60
11.93
2.37
0.09
0.13
1448
148
99.88
DSDP051-0417D-022-007/28-30
144.78
50.42
1.61
14.47
11.27
0.20
7.63
11.82
2.49
0.08
0.14
1450
114
100.50
Table A2: Major element composition of ancient Atlantic MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP051-0417D-026-001/50-54
150.00
50.11
1.58
14.53
11.11
0.22
7.68
11.96
2.47
0.08
0.11
1460
547
100.27
DSDP051-0417D-026-002/42-44
151.22
50.23
1.64
14.37
11.31
0.21
7.61
11.92
2.44
0.09
0.13
1556
73
100.34
DSDP051-0417D-026-006/47-51
156.87
50.29
1.53
14.73
10.96
0.23
7.78
11.94
2.44
0.08
0.13
1393
349
100.50
DSDP051-0417D-026-007/3-5
157.79
49.69
1.53
14.72
10.91
0.21
7.74
11.93
2.48
0.08
0.12
1370
368
99.79
DSDP051-0417D-027-002/5-7
160.06
49.98
1.52
14.71
10.87
0.20
7.86
11.88
2.44
0.08
0.11
1443
156
100.03
DSDP051-0417D-027-003/27-29
161.64
50.51
1.55
14.68
11.00
0.22
7.81
11.82
2.41
0.09
0.12
1388
239
100.58
DSDP051-0417D-027-004/51-53
163.21
50.10
1.54
14.73
10.99
0.21
7.83
11.88
2.46
0.09
0.13
1400
96
100.31
DSDP051-0417D-027-005/56-59
164.60
49.83
1.52
14.72
10.88
0.20
7.83
11.91
2.48
0.09
0.11
1322
239
99.91
DSDP051-0417D-027-006/123-125 166.57
50.25
1.53
14.76
10.88
0.20
7.93
11.91
2.43
0.10
0.11
1417
180
100.46
DSDP051-0417D-027-007/3-7
167.00
49.86
1.66
14.56
11.33
0.22
7.57
11.80
2.50
0.09
0.13
1534
244
100.14
DSDP051-0417D-028-001/104-105 168.74
50.01
1.52
14.75
10.90
0.20
7.84
11.99
2.46
0.09
0.12
1421
385
100.29
DSDP051-0417D-028-002/16-18
169.22
50.00
1.53
14.73
10.86
0.20
7.87
11.93
2.43
0.08
0.10
1451
190
100.11
DSDP051-0417D-028-002/85-87
169.91
49.87
1.52
14.73
10.83
0.22
7.87
11.94
2.48
0.09
0.11
1384
173
100.02
DSDP051-0417D-028-003/53-57
170.97
50.01
1.51
14.82
10.87
0.22
7.91
11.94
2.44
0.09
0.12
1410
79
100.28
DSDP051-0417D-028-004/63-64
172.43
50.41
1.52
14.79
10.87
0.22
7.85
12.03
2.43
0.08
0.13
1383
248
100.70
DSDP051-0417D-028-006/8-10
174.66
50.08
1.50
14.79
10.87
0.20
7.89
11.94
2.46
0.09
0.13
1442
70
100.30
DSDP051-0417D-029-001/108-110
177.88
50.09
1.52
14.72
10.81
0.20
7.87
11.95
2.45
0.09
0.11
1371
95
100.18
DSDP051-0417D-029-002/126-129 179.53
50.48
1.52
14.83
10.88
0.20
7.81
11.92
2.45
0.08
0.12
1485
229
100.66
DSDP051-0417D-029-003/120-122 180.84
50.02
1.51
14.79
10.84
0.21
7.95
11.98
2.47
0.08
0.10
1372
356
100.33
DSDP051-0417D-029-004/54-56
181.63
50.13
1.51
14.81
10.88
0.20
7.88
11.94
2.46
0.08
0.13
1434
76
100.37
DSDP051-0417D-029-006/95-100
184.79
50.45
1.53
14.73
10.90
0.20
7.86
11.91
2.44
0.08
0.12
1398
196
100.58
DSDP051-0417D-030-001/88-91
186.78
50.32
1.52
14.70
10.89
0.20
7.79
11.94
2.41
0.08
0.13
1335
164
100.33
DSDP051-0417D-030-002/99-100
188.09
50.19
1.54
14.74
10.92
0.20
7.83
12.02
2.44
0.08
0.13
1416
124
100.46
DSDP051-0417D-030-003/129-132 189.69
50.37
1.55
14.70
10.92
0.20
7.92
11.97
2.46
0.08
0.13
1420
252
100.66
DSDP051-0417D-030-004/54-56
190.33
50.22
1.54
14.72
10.98
0.19
7.79
11.95
2.47
0.09
0.12
1431
100
100.43
DSDP051-0417D-030-005/94-96
192.08
50.25
1.52
14.70
11.00
0.21
7.85
11.96
2.47
0.08
0.11
1482
299
100.56
DSDP051-0417D-030-006/36-38
192.85
50.26
1.56
14.64
11.09
0.21
7.71
11.94
2.49
0.08
0.12
1416
230
100.47
DSDP051-0417D-031-001/11-15
195.11
50.38
1.55
14.56
11.02
0.19
7.72
11.90
2.41
0.08
0.14
1420
253
100.34
DSDP051-0417D-031-002/129-131 197.74
50.29
1.58
14.60
11.11
0.22
7.72
11.90
2.45
0.09
0.12
1498
133
100.46
DSDP051-0417D-031-004/17-18
199.54
50.10
1.57
14.61
10.98
0.22
7.80
11.84
2.46
0.09
0.13
1386
232
100.16
DSDP051-0417D-034-005/118-120
227.58
50.19
1.67
14.29
11.37
0.21
7.41
11.88
2.47
0.09
0.13
1598
347
100.15
DSDP051-0417D-035-005/126-128 236.58
49.31
1.58
14.76
11.26
0.21
7.33
11.97
2.35
0.09
0.14
1437
88
99.36
DSDP051-0417D-037-001/34-36
243.34
49.86
1.58
14.64
11.29
0.20
7.39
11.98
2.36
0.08
0.13
1374
358
99.89
DSDP051-0417D-037-004/33-36
247.39
49.75
1.58
14.82
11.30
0.22
7.29
11.95
2.37
0.09
0.11
1435
68
99.83
DSDP051-0417D-037-007/25-28
251.48
49.86
1.60
14.65
11.33
0.21
7.41
11.95
2.38
0.09
0.12
1519
269
100.00
DSDP051-0417D-038-002/4-6
251.80
49.82
1.60
14.68
11.34
0.22
7.41
11.94
2.35
0.09
0.14
1518
296
99.99
DSDP051-0417D-039-006/57-58
263.79
50.20
1.50
14.64
11.09
0.20
7.46
12.14
2.33
0.08
0.14
1389
322
100.17
DSDP051-0417D-040-003/7-9
265.95
50.13
1.54
14.62
11.11
0.20
7.42
12.15
2.36
0.09
0.12
1486
237
100.14
DSDP051-0417D-041-004/107-109 273.01
48.85
1.39
14.44
10.22
0.18
7.50
11.88
2.28
0.08
0.11
1374
467
97.32
DSDP051-0417D-042-002/8-12
278.25
49.83
1.41
14.89
10.72
0.22
7.58
12.23
2.27
0.08
0.11
1390
210
99.69
DSDP051-0417D-044-004/7-9
299.22
49.83
1.65
14.30
11.41
0.22
7.16
11.83
2.39
0.09
0.14
1543
211
99.42
DSDP052-0417D-055-002/29-31
380.19
50.23
1.70
14.38
11.76
0.22
7.12
11.63
2.41
0.08
0.13
1601
394
100.08
DSDP052-0417D-060-006/39-40
432.16
49.91
1.49
15.04
10.91
0.21
7.54
12.03
2.37
0.09
0.12
1385
159
100.07
Table A2: Major element composition of ancient Atlantic MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP052-0417D-062-003/12-13
437.02
49.85
1.60
14.86
11.27
0.21
7.28
11.82
2.37
0.10
0.14
1495
237
99.90
DSDP052-0417D-062-006/81-83
442.06
50.14
1.59
14.83
11.35
0.20
7.37
11.76
2.39
0.10
0.13
1480
308
100.26
DSDP052-0417D-063-004/133-135 448.66
50.02
1.59
14.86
11.34
0.20
7.37
11.77
2.39
0.09
0.13
1435
120
100.13
DSDP052-0417D-063-006/79-80
451.02
49.93
1.59
14.93
11.29
0.19
7.29
11.74
2.40
0.10
0.12
1419
171
99.95
DSDP052-0417D-064-002/128-130 454.74
50.40
1.60
14.75
11.29
0.22
7.44
11.76
2.42
0.09
0.13
1443
158
100.49
DSDP052-0417D-064-004/40-43
456.67
50.25
1.58
14.71
11.32
0.22
7.41
11.81
2.41
0.10
0.14
1495
104
100.33
DSDP052-0417D-066-002/28-30
471.75
50.17
1.55
14.94
11.23
0.22
7.42
11.74
2.40
0.10
0.11
1450
263
100.26
DSDP052-0417D-066-004/80-83
474.70
50.27
1.56
14.84
11.17
0.21
7.41
11.81
2.42
0.09
0.12
1412
153
100.26
DSDP052-0417D-066-005/33-34
475.70
50.36
1.54
14.79
11.11
0.21
7.46
11.78
2.43
0.09
0.12
1400
243
100.27
DSDP052-0417D-066-006/41-43
477.22
49.26
1.50
14.77
10.59
0.20
7.31
11.46
2.33
0.09
0.10
1392
130
97.97
DSDP052-0418A-015-001/64-66
1.40
51.13
1.24
14.43
10.02
0.21
8.03
12.06
2.20
0.08
0.09
1238
35
99.81
DSDP052-0418A-015-001/140-142
0.64
50.94
1.22
14.34
10.04
0.21
8.03
12.15
2.12
0.08
0.11
1264
25
99.56
DSDP052-0418A-043-001/15-17
196.65
49.69
1.12
15.26
9.58
0.19
8.56
12.37
2.04
0.04
0.10
1148
29
99.24
DSDP052-0418A-045-001/100-102
215.50
50.06
1.19
14.99
9.88
0.20
8.38
12.24
2.04
0.06
0.08
1216
32
99.42
DSDP052-0418A-048-001/56-58
242.06
50.09
1.22
14.94
10.04
0.20
8.35
12.03
2.15
0.05
0.09
1251
9
99.48
DSDP053-0418A-055-001/5-6
300.45
50.94
1.56
13.72
11.32
0.23
7.29
11.61
2.17
0.08
0.13
1435
43
99.41
DSDP053-0418A-059-004/3-4
341.63
50.49
1.68
14.10
11.44
0.22
7.22
11.33
2.26
0.11
0.14
1500
48
99.37
DSDP053-0418A-064-003/53-55
378.98
51.08
1.39
14.17
10.53
0.21
7.75
11.96
2.20
0.08
0.11
1353
43
99.84
DSDP053-0418A-068-001/133-134
410.43
50.97
1.56
13.92
11.19
0.22
7.50
11.63
2.22
0.08
0.12
1451
16
99.79
DSDP053-0418A-072-001/50-52
439.30
51.04
1.53
13.97
11.12
0.22
7.47
11.74
2.20
0.08
0.14
1396
17
99.85
DSDP053-0418A-074-004/44-45
459.08
50.92
1.52
14.02
11.01
0.21
7.33
11.65
2.14
0.09
0.12
1436
33
99.36
DSDP053-0418A-075-005/3-6
467.29
51.46
1.42
14.06
10.86
0.21
7.58
11.62
2.23
0.11
0.13
1345
45
100.01
DSDP053-0418A-086-004/32-34
544.14
50.70
1.37
14.49
10.46
0.18
7.82
12.00
2.03
0.08
0.12
1290
46
99.58
DSDP073-0522B-003-002/117-119
0.57
50.28
1.55
14.79
9.45
0.20
7.76
12.76
2.74
0.13
0.13
1218
60
100.10
DSDP073-0522B-003-002/128-134
0.80
50.28
1.57
14.80
9.41
0.20
7.71
12.77
2.70
0.13
0.12
1269
51
100.01
DSDP073-0522B-003-003/75-78
2.07
50.01
1.56
14.79
9.43
0.19
7.71
12.65
2.73
0.13
0.13
1149
52
99.62
DSDP073-0522B-003-003/117-121
1.65
50.13
1.52
14.68
9.08
0.18
7.70
12.55
2.65
0.13
0.13
1200
78
99.05
DSDP073-0522B-003-004/4-7
2.39
50.26
1.57
15.08
9.55
0.20
7.76
12.82
2.21
0.19
0.13
1130
211
100.07
DSDP073-0522B-004-001/82-85
5.82
50.66
1.56
15.35
8.59
0.17
7.26
12.80
3.22
0.18
0.14
185
165
100.00
DSDP073-0522B-004-002/56-57
6.73
51.11
1.58
14.92
9.63
0.19
7.66
12.82
2.41
0.17
0.12
1218
94
100.92
DSDP073-0522B-005-001/11-13
9.51
50.03
1.52
14.83
9.50
0.18
7.68
12.71
2.71
0.12
0.16
1195
37
99.74
DSDP073-0522B-005-001/113-115
10.85
49.49
1.50
15.03
9.34
0.18
7.76
12.63
2.72
0.13
0.14
1187
87
99.23
DSDP073-0522B-005-002/6-7
10.91
49.31
1.52
15.05
9.39
0.19
7.73
12.68
2.66
0.13
0.13
1207
84
99.10
DSDP073-0522B-005-003/139-141
12.22
49.98
1.54
14.81
9.42
0.18
7.67
12.70
2.67
0.13
0.13
1190
53
99.53
DSDP073-0522B-006-001/10-12
16.50
50.59
1.36
14.52
9.85
0.19
7.86
12.82
2.48
0.05
0.09
1228
42
100.12
DSDP076-0534A-128-004/136-139
10.36
50.99
0.91
14.21
10.74
0.21
7.45
12.27
1.82
0.10
0.07
1184
68
99.07
DSDP076-0534A-129-005/6-8
19.36
51.57
0.93
14.23
10.85
0.21
7.67
12.26
1.86
0.10
0.08
1094
60
100.03
DSDP076-0534A-130-001/72-74
23.50
50.74
0.95
14.18
10.80
0.21
7.70
12.18
1.85
0.10
0.08
1138
71
99.08
DSDP076-0534A-130-001/100-101
23.22
51.82
0.94
14.17
10.80
0.20
7.64
12.22
1.85
0.10
0.07
1118
67
100.10
DSDP078-0543A-013-001/10-11
16.10
49.32
1.59
14.83
9.65
0.21
7.60
11.88
2.68
0.12
0.15
1252
53
98.36
DSDP078-0543A-013-002/2-3
18.55
49.25
1.59
14.89
9.73
0.19
7.48
11.80
2.69
0.12
0.12
1329
33
98.20
DSDP078-0543A-013-002/105-106
17.52
48.78
1.64
15.00
9.75
0.18
7.51
11.78
2.71
0.13
0.16
1301
36
97.97
DSDP078-0543A-013-004/125-126
21.49
49.07
1.59
14.99
9.72
0.19
7.56
11.83
2.65
0.12
0.17
1179
32
98.18
Table A2: Major element composition of ancient Atlantic MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP078-0543A-013-005/98-99
22.72
48.66
1.57
14.80
9.66
0.18
7.62
11.93
2.70
0.12
0.16
1316
22
97.75
DSDP078-0543A-014-001/57-59
25.57
49.29
1.59
14.88
9.72
0.19
7.54
11.78
2.74
0.13
0.16
1265
45
98.34
DSDP078-0543A-015-002/2-3
28.52
50.30
1.67
14.89
9.91
0.19
7.35
11.65
2.80
0.13
0.17
1361
41
99.41
DSDP078-0543A-015-003/137-139
31.37
50.95
1.56
14.85
9.64
0.18
7.55
11.95
2.67
0.12
0.14
1217
26
99.92
DSDP078-0543A-015-005/18-21
33.18
49.11
1.57
14.85
9.69
0.18
7.62
11.95
2.72
0.12
0.16
1212
22
98.27
DSDP078-0543A-016-001/139-140
35.39
49.98
1.59
14.77
9.61
0.18
7.67
11.93
2.72
0.12
0.13
1278
46
99.02
DSDP078-0543A-016-003/57-59
37.52
50.06
1.74
14.89
9.89
0.19
7.37
11.63
2.81
0.13
0.18
1367
52
99.23
DSDP078-0543A-016-004/116-117
39.48
50.53
1.70
14.94
9.87
0.20
7.31
11.57
2.78
0.13
0.17
1314
60
99.52
DSDP078-0543A-016-006/2-3
42.20
49.24
1.79
14.81
10.06
0.21
7.34
11.46
2.89
0.14
0.18
1289
59
98.43
DSDP078-0543A-016-006/118-120
41.04
49.17
1.76
14.91
10.11
0.20
7.40
11.67
2.82
0.14
0.16
1366
25
98.68
DSDP078-0543A-016-007/105-106
43.42
50.64
1.66
14.98
9.76
0.21
7.49
11.72
2.77
0.12
0.17
1333
53
99.86
DiB
Depth in basement [m]
Table A3: Major element composition of ancient Pacific MORB (uncorrected raw data).
sample ID
DiB
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[m]
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP016-0163-028-001/113-114
1.13
48.69
1.13
15.22
8.96
0.17
8.60
12.09
2.35
0.05
0.08
1121
43
97.63
DSDP016-0163-029-001/90-92
9.90
49.70
1.18
15.57
9.56
0.18
8.79
12.56
2.37
0.05
0.10
1183
44
100.35
DSDP017-0166-029-001/82-101
0.82
49.31
2.29
13.37
14.33
0.22
6.34
11.25
2.84
0.10
0.19
1828
672
100.77
DSDP017-0166-029-002/122-123
2.72
49.78
2.25
13.23
14.24
0.23
6.38
11.21
2.87
0.10
0.22
1825
716
101.03
DSDP029-0278-035-001/78-80
1.78
50.70
1.03
15.67
8.38
0.17
8.80
13.34
2.30
0.09
0.10
1060
63
100.84
DSDP029-0278-035-001/111-112
2.11
50.47
1.03
15.64
8.40
0.15
8.89
13.27
2.33
0.09
0.10
1002
75
100.61
DSDP029-0278-035-002/90-92
3.40
50.54
1.04
15.73
8.45
0.16
8.77
13.29
2.28
0.09
0.09
987
97
100.70
DSDP029-0278-035-003/10-12
4.10
51.12
1.02
15.61
8.48
0.16
8.95
13.19
2.30
0.09
0.10
1070
75
101.29
DSDP029-0278-035-003/68-69
4.68
51.00
1.02
15.62
8.44
0.14
8.92
13.17
2.34
0.09
0.09
1011
65
101.09
DSDP063-0469-044-002/91-92
7.61
50.35
1.97
13.87
12.10
0.24
6.91
11.47
2.60
0.10
0.17
1523
200
100.18
DSDP063-0469-048-001/54-57
32.74
50.88
1.71
14.19
11.68
0.21
7.29
11.86
2.48
0.07
0.13
1550
175
100.90
DSDP063-0469-048-001/115-116
33.35
49.54
1.86
14.03
12.02
0.22
7.11
11.50
2.54
0.07
0.13
1376
113
99.39
DSDP063-0469-049-001/106-107
42.26
50.64
1.78
14.02
11.85
0.23
7.32
11.67
2.57
0.08
0.13
1456
109
100.66
DSDP063-0469-050-001/80-81
50.00
50.42
1.80
14.26
11.90
0.22
7.17
11.42
2.63
0.08
0.15
1493
182
100.43
DSDP063-0469-050-001/111-112
50.31
50.30
1.80
14.23
11.96
0.22
7.24
11.34
2.63
0.08
0.12
1518
110
100.30
DSDP063-0469-050-002/11-12
50.81
50.30
1.78
14.14
11.96
0.22
7.28
11.42
2.66
0.08
0.13
1471
123
100.35
DSDP063-0470A-007-001/67-68
0.17
50.16
1.57
15.14
10.17
0.21
7.95
12.42
2.31
0.09
0.15
1178
104
100.47
DSDP063-0470A-007-002/23-24
1.23
48.93
1.23
16.55
8.68
0.16
8.72
12.58
2.21
0.19
0.10
969
219
99.60
DSDP063-0470A-007-003/53-54
3.03
49.29
1.25
16.53
8.58
0.15
8.70
12.58
2.26
0.23
0.14
954
209
99.97
DSDP063-0470A-008-001/59-60
4.59
50.00
1.37
15.91
9.42
0.17
8.28
12.57
2.66
0.03
0.10
1144
156
100.81
DSDP063-0470A-008-002/26-27
5.76
49.76
1.36
15.98
9.35
0.16
8.27
12.59
2.66
0.03
0.11
1198
114
100.58
DSDP063-0470A-008-003/112-113
8.12
48.67
1.35
15.98
9.38
0.17
8.20
12.51
2.62
0.03
0.13
1223
118
99.36
DSDP063-0470A-008-004/142-143
9.92
49.06
1.36
15.89
9.41
0.16
8.33
12.52
2.71
0.04
0.10
1204
128
99.90
DSDP063-0470A-008-005/63-65
10.58
49.26
1.37
15.94
9.43
0.18
8.35
12.53
2.68
0.04
0.11
1260
135
100.21
DSDP063-0470A-009-001/81-82
14.31
49.38
1.36
16.00
9.39
0.18
8.28
12.53
2.63
0.03
0.10
1178
117
100.21
DSDP063-0470A-011-001/3-4
27.53
49.77
1.39
15.95
9.47
0.18
8.16
12.61
2.66
0.04
0.12
1173
163
100.65
DSDP063-0470A-011-001/64-65
28.14
50.65
1.45
14.87
10.50
0.20
7.63
12.00
2.56
0.08
0.12
1217
58
100.38
DSDP063-0470A-012-001/34-35
31.84
48.76
1.90
15.88
9.89
0.19
7.57
11.38
2.75
0.33
0.23
1187
160
99.20
DSDP063-0470A-013-001/40-41
40.90
50.72
1.83
14.63
11.03
0.19
7.43
11.81
2.52
0.11
0.17
1361
276
100.82
DSDP063-0472-014-001/13-14
0.13
49.57
1.24
15.22
8.72
0.17
8.36
12.05
2.57
0.17
0.11
1124
145
98.49
DSDP068-0501-010-001/49-50
0.49
50.65
1.12
13.87
10.31
0.19
7.83
12.29
2.17
0.02
0.09
1418
91
98.90
DSDP068-0501-014-003/138-140
23.23
50.42
0.98
14.91
9.19
0.18
8.88
12.91
2.00
0.01
0.08
1220
33
99.86
DSDP068-0501-014-004/61-63
23.96
50.84
0.96
15.07
9.27
0.16
8.80
13.02
2.00
0.02
0.04
1145
43
100.47
DSDP068-0501-015-001/76-77
28.76
51.14
0.94
15.03
9.27
0.17
8.82
12.97
2.01
0.02
0.06
1169
34
100.72
DSDP068-0501-015-003/64-65
31.64
50.69
0.97
15.14
9.21
0.18
8.90
13.04
2.00
0.02
0.05
1177
16
100.48
DSDP068-0501-015-004/2-3
32.52
50.50
0.97
15.10
9.20
0.18
8.78
13.04
2.04
0.02
0.04
1139
25
100.16
DSDP068-0501-017-002/113-115
43.43
50.15
0.97
15.21
9.24
0.18
8.59
13.07
1.98
0.02
0.07
1115
32
99.77
DSDP068-0501-020-003/121-122
68.21
50.80
1.03
14.30
10.23
0.20
8.03
12.77
2.05
0.01
0.05
1334
113
99.82
DSDP069-0504B-004-001/66-67
6.16
48.05
0.97
15.40
9.29
0.18
8.87
12.92
1.94
0.02
0.06
1114
29
97.98
DSDP069-0504B-005-002/78-79
17.28
51.44
0.97
15.25
9.31
0.19
8.72
13.13
1.96
0.02
0.07
1123
36
101.35
DSDP069-0504B-006-002/142-144
26.92
51.63
0.99
15.33
9.27
0.19
8.88
13.11
1.98
0.02
0.07
1157
23
101.75
DSDP069-0504B-012-002/38-40
79.88
50.82
0.99
15.18
9.14
0.19
8.71
13.01
2.16
0.02
0.07
1190
54
100.57
DSDP069-0504B-015-002/51-52
102.51
51.73
0.96
14.92
9.77
0.19
8.67
13.02
1.93
0.02
0.06
1196
16
101.57
Table A3: Major element composition of ancient Pacific MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
DSDP069-0504B-015-005/64-66
107.14
50.55
0.95
14.98
9.79
0.19
8.60
13.02
1.91
0.01
0.05
1199
26
100.35
DSDP069-0504B-016-003/123-124
113.73
50.53
1.08
15.04
9.28
0.19
8.36
13.05
2.05
0.04
0.09
1167
73
100.00
DSDP069-0504B-018-001/10-12
123.60
49.85
1.08
15.35
9.37
0.18
8.69
13.22
2.23
0.03
0.07
1234
69
100.37
DSDP069-0504B-021-004/108-109
152.08
50.88
0.88
14.93
9.38
0.18
8.83
13.44
1.88
0.02
0.05
1150
27
100.75
DSDP069-0504B-022-002/29-30
157.29
50.70
0.86
14.92
9.29
0.16
8.81
13.30
1.86
0.02
0.05
1079
22
100.23
DSDP069-0504B-025-001/102-104
183.52
51.59
1.05
14.69
9.76
0.19
8.35
13.11
2.18
0.01
0.06
1261
35
101.31
DSDP069-0504B-026-001/93-94
189.43
51.34
1.06
14.70
9.70
0.17
8.33
13.13
2.09
0.02
0.07
1260
36
100.92
DSDP069-0504B-028-003/42-43
203.92
51.87
1.10
14.56
10.05
0.19
8.35
12.88
2.11
0.02
0.07
1298
39
101.52
DSDP069-0504B-029-002/111-112
212.11
51.61
1.05
14.75
9.62
0.19
8.43
13.27
2.15
0.01
0.07
1175
57
101.46
DSDP070-0504B-035-001/17-19
260.17
49.62
0.82
17.09
8.47
0.17
9.79
13.21
2.20
0.02
0.04
1017
45
101.68
DSDP070-0504B-038-001/6-7
287.06
51.12
0.96
14.84
9.79
0.18
8.51
13.20
2.10
0.02
0.05
1200
40
101.08
DSDP070-0504B-045-002/23-24
347.23
52.51
1.14
14.38
10.67
0.20
7.97
12.51
2.08
0.02
0.08
1270
28
101.87
DSDP070-0504B-048-002/123-124
375.23
51.82
0.98
14.65
9.77
0.19
8.37
13.07
1.90
0.01
0.06
1207
22
101.14
DSDP070-0504B-049-003/47-48
384.87
51.39
0.95
14.74
9.77
0.19
8.47
13.10
1.88
0.01
0.05
1159
35
100.85
DSDP070-0504B-056-001/87-88
431.87
50.25
1.39
15.21
9.68
0.21
8.22
12.65
2.46
0.08
0.15
1153
40
100.58
DSDP070-0504B-061-001/31-32
476.31
50.04
1.08
15.11
9.98
0.20
8.66
12.88
1.89
0.02
0.07
1123
32
100.20
DSDP070-0504B-064-001/29-30
498.79
51.31
1.04
14.86
9.60
0.17
8.56
13.05
1.98
0.01
0.07
1198
16
100.97
DSDP085-0573B-043-001/27-28
0.27
50.48
1.97
14.21
11.41
0.21
7.20
11.72
2.89
0.09
0.15
1544
358
100.76
DSDP091-0595B-007-002/67-70
47.07
50.77
1.71
14.14
11.19
0.21
7.37
11.72
2.73
0.10
0.17
1427
280
100.50
DSDP091-0595B-007-002/80-81
47.20
50.20
1.77
14.28
11.24
0.21
7.34
11.71
2.77
0.10
0.16
1496
296
100.17
DSDP092-0597A-007-CC/5-6
0.05
50.91
0.90
14.90
9.60
0.16
8.77
13.23
1.97
0.04
0.06
1130
99
100.82
ODP0185-0801C-020-003/31-32
170.23
49.23
2.11
13.37
12.64
0.23
6.83
11.03
2.72
0.13
0.19
1900
689
99.02
ODP0185-0801C-023-003/130-132
192.18
48.90
2.17
13.39
12.69
0.21
6.51
10.65
2.70
0.15
0.17
1764
618
98.03
ODP0185-0801C-027-003/2-3
228.80
49.43
1.83
14.01
11.58
0.22
7.30
11.59
2.55
0.14
0.17
1593
339
99.25
ODP0185-0801C-032-001/68-70
273.28
49.11
2.42
13.29
13.13
0.25
6.53
10.66
2.35
0.13
0.21
1903
690
98.63
ODP0185-0801C-034-001/123-124
292.83
48.64
2.41
13.27
13.17
0.24
6.28
10.46
2.23
0.14
0.21
1931
816
97.62
ODP0185-0801C-035-004/63-74
306.05
48.33
2.37
13.06
13.13
0.26
6.39
10.44
2.75
0.15
0.21
1980
877
97.67
ODP0185-0801C-042-002/73-89
368.24
48.99
1.39
15.40
10.26
0.19
8.19
12.26
2.46
0.08
0.12
1442
158
99.72
ODP0185-0801C-042-002/90-94
368.41
49.16
1.32
14.96
9.86
0.18
8.40
12.12
2.50
0.08
0.11
1347
161
99.06
ODP0185-0801C-042-002/94-100
368.45
49.86
1.37
15.11
10.17
0.20
8.34
12.42
2.52
0.09
0.13
1300
148
100.55
ODP0185-1149D-009-002/71-72
40.23
50.94
1.78
14.29
11.60
0.23
7.23
11.76
2.40
0.14
0.16
1570
182
100.94
ODP0191-1179D-019-003/126-127
63.78
51.37
1.59
13.84
11.89
0.21
7.31
11.51
2.45
0.10
0.13
1493
456
100.83
ODP0191-1179D-020-002/97-99
71.82
49.62
1.50
15.92
9.87
0.18
8.64
11.88
2.51
0.16
0.15
1204
91
100.73
ODP0191-1179D-020-005/56-60
75.33
49.54
1.53
15.99
9.90
0.18
8.72
11.75
2.50
0.16
0.15
1196
70
100.72
ODP0191-1179D-021-001/94-95
80.14
49.40
1.53
16.16
9.95
0.18
8.70
11.75
2.52
0.16
0.17
1169
96
100.81
ODP0191-1179D-021-002/52-53
80.98
49.54
1.53
16.09
9.95
0.17
8.81
11.84
2.50
0.18
0.17
1172
59
101.06
ODP0191-1179D-021-003/38-41
82.09
49.71
1.55
16.11
9.92
0.17
8.82
11.77
2.59
0.18
0.17
1213
76
101.28
ODP0191-1179D-021-004/64-67
83.85
49.63
1.56
16.22
9.90
0.19
8.61
11.76
2.56
0.17
0.15
1207
124
101.08
ODP0191-1179D-022-001/91-92
89.71
49.36
1.58
16.34
9.89
0.19
8.55
11.83
2.52
0.18
0.17
1166
79
100.91
ODP0191-1179D-022-001/104-105
89.84
49.64
1.56
16.25
9.76
0.18
8.48
11.64
2.54
0.19
0.19
1186
91
100.73
ODP0191-1179D-022-003/46-51
92.14
50.33
1.53
15.61
10.20
0.19
8.47
11.87
2.50
0.15
0.14
1298
93
101.34
ODP0191-1179D-022-003/55-56
92.23
49.75
1.55
16.10
9.91
0.20
8.95
11.68
2.57
0.18
0.16
1233
67
101.34
ODP0199-1215B-011-CC/4-6
19.24
49.47
2.03
13.48
13.15
0.23
6.55
11.20
2.72
0.08
0.19
1648
257
99.54
Table A3: Major element composition of ancient Pacific MORB (uncorrected raw data).
sample ID
DiB
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[m]
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
ODP0199-1217A-017-CC/1-8
0.01
51.27
1.43
14.05
10.87
0.20
7.20
12.08
2.73
0.07
0.12
1299
26
100.33
ODP0199-1217A-017-CC/9-13
0.09
49.98
1.43
14.05
10.80
0.17
7.22
12.16
2.68
0.07
0.14
1279
24
99.04
ODP0199-1222A-012-CC/9-10
0.09
49.96
2.45
13.17
13.67
0.22
5.96
10.50
2.87
0.18
0.27
1718
495
99.72
ODP0199-1222A-012-CC/30-32
0.21
49.82
2.48
13.60
13.71
0.24
6.25
10.56
2.39
0.18
0.25
1744
453
99.95
ODP0200-1224F-004-006/16-17
34.64
50.39
2.12
13.74
13.25
0.24
6.65
11.54
1.94
0.10
0.21
1640
290
100.62
ODP0200-1224F-006-001/28
47.28
48.95
2.00
13.62
13.03
0.23
6.57
11.39
2.62
0.10
0.20
1642
307
99.14
ODP0200-1224F-007-001/101-102
57.51
48.89
2.03
13.63
12.86
0.20
6.62
11.34
2.62
0.09
0.20
1611
322
98.91
ODP0200-1224F-008-001/73-74
66.23
49.47
2.04
13.69
13.13
0.21
6.73
11.61
1.99
0.11
0.21
1759
321
99.66
ODP0200-1224F-009-001/6-7
74.76
49.44
2.02
13.31
12.93
0.22
6.59
11.39
2.63
0.11
0.19
1632
328
99.27
ODP0200-1224F-010-001/40-46
84.20
49.25
2.03
13.52
12.94
0.18
6.67
11.32
2.55
0.10
0.18
1640
320
99.20
ODP0200-1224F-011-002/12-18
94.72
49.78
2.42
12.74
14.26
0.25
5.90
10.62
2.85
0.13
0.24
1881
627
99.72
ODP0200-1224F-012-001/61-70
101.11
47.56
3.17
12.24
16.39
0.28
5.25
9.65
2.81
0.16
0.33
2172
772
98.45
ODP0203-1243B-008-001/28-40
33.68
48.78
2.11
16.06
8.79
0.17
6.59
10.02
3.61
0.71
0.36
1158
605
97.54
ODP0203-1243B-013-001/88-90
58.38
50.58
2.88
13.95
12.11
0.22
5.29
9.35
2.75
0.62
0.38
1723
712
98.64
ODP0206-1256C-007-002/111-112
17.09
50.62
1.89
13.31
12.95
0.24
6.46
10.73
2.64
0.12
0.15
1731
878
99.63
ODP0206-1256C-007-003/101-102
18.47
50.85
1.89
13.45
13.03
0.22
6.40
10.84
2.34
0.22
0.13
1712
873
99.88
ODP0206-1256C-007-004/45-49
19.06
50.52
1.81
13.47
12.63
0.22
6.40
10.67
2.58
0.13
0.15
1696
884
99.09
ODP0206-1256C-007-005/116-124
20.27
50.37
1.85
13.30
12.70
0.22
6.49
10.47
2.60
0.13
0.17
1693
884
98.79
ODP0206-1256C-008-004/130-142
29.09
51.58
1.91
13.48
12.92
0.24
6.57
10.70
2.28
0.14
0.15
1762
960
100.49
ODP0206-1256C-008-005/8-18
29.37
51.13
1.92
13.60
13.21
0.24
6.51
10.87
2.51
0.14
0.17
1658
903
100.80
ODP0206-1256D-013-002/133-134
103.10
50.28
1.77
14.18
10.67
0.20
7.09
11.77
2.68
0.25
0.20
1317
694
99.49
ODP0206-1256D-014-002/33-37
111.33
50.52
1.81
14.26
10.86
0.22
7.03
11.88
2.71
0.26
0.19
1397
696
100.13
ODP0206-1256D-014-002/84-95
111.84
49.87
1.76
14.21
10.70
0.20
7.00
11.70
2.63
0.25
0.22
1325
718
98.95
ODP0206-1256D-014-003/1-2
112.51
50.03
1.78
14.17
10.73
0.21
7.00
11.76
2.70
0.25
0.21
1374
705
99.24
ODP0206-1256D-015-001/5-15
114.25
49.69
1.77
14.28
10.72
0.21
7.15
11.79
2.75
0.25
0.20
1438
732
99.23
ODP0206-1256D-015-002/27-28
115.97
52.00
1.04
13.80
10.94
0.19
7.58
11.92
2.03
0.09
0.09
1238
558
100.04
ODP0206-1256D-015-003/2-3
117.22
51.84
1.04
13.84
10.86
0.20
7.35
11.87
1.98
0.09
0.10
1193
529
99.51
ODP0206-1256D-016-001/119-120
120.09
52.07
1.25
13.51
11.87
0.22
7.17
11.47
2.25
0.08
0.09
1393
507
100.37
ODP0206-1256D-017-001/63-67
124.13
51.69
1.29
13.55
11.84
0.23
7.07
11.40
2.19
0.08
0.09
1383
491
99.83
ODP0206-1256D-018-001/75-80
128.75
51.83
1.27
13.52
11.78
0.20
6.97
11.42
2.15
0.08
0.10
1286
553
99.70
ODP0206-1256D-018-002/56-67
129.82
51.81
1.24
13.52
11.77
0.21
7.13
11.45
2.22
0.08
0.10
1390
471
99.92
ODP0206-1256D-020-001/36-37
137.76
51.86
1.38
13.50
12.14
0.22
6.84
11.18
2.24
0.08
0.10
1449
516
99.96
ODP0206-1256D-021-001/33-35
147.13
52.37
1.53
13.24
12.82
0.22
5.86
10.42
2.46
0.15
0.20
1473
1432
99.78
ODP0206-1256D-021-001/87-88
147.67
54.50
1.95
12.73
14.02
0.24
4.23
9.07
2.59
0.27
0.36
1583
3543
100.70
ODP0206-1256D-021-002/1-6
148.26
53.62
2.03
12.38
14.21
0.24
3.70
8.52
2.82
0.30
0.38
1439
3744
98.94
ODP0206-1256D-021-002/12-18
148.37
54.58
1.93
12.82
14.00
0.26
4.41
9.22
2.53
0.26
0.35
1503
3373
101.09
ODP0206-1256D-021-002/36-49
148.61
54.03
1.97
12.77
13.86
0.24
4.23
9.08
2.71
0.27
0.35
1470
3435
100.22
ODP0206-1256D-022-004/86-87
160.88
51.76
1.25
13.62
11.64
0.22
7.14
11.61
2.17
0.07
0.08
1349
467
99.94
ODP0206-1256D-023-002/16-20
161.85
51.72
1.25
13.69
11.63
0.20
7.09
11.60
2.13
0.08
0.09
1213
423
99.82
ODP0206-1256D-024-001/83-92
170.33
51.65
1.12
13.96
10.67
0.21
7.52
12.07
2.04
0.08
0.11
1192
465
99.78
ODP0206-1256D-024-002/20-25
171.16
51.50
1.11
13.96
10.65
0.19
7.52
12.09
2.03
0.08
0.09
1224
486
99.58
ODP0206-1256D-024-002/102-106
171.98
51.83
1.08
13.92
10.69
0.20
7.61
12.11
2.02
0.08
0.09
1172
556
99.97
ODP0206-1256D-026-001/146-147
189.76
51.02
1.38
13.40
12.07
0.20
6.70
11.05
2.20
0.64
0.10
1399
533
99.17
Table A3: Major element composition of ancient Pacific MORB (uncorrected raw data).
sample ID
DiB
[m]
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K 2O
P2O5
S
Cl
Total
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [ppm] [ppm] [wt. %]
ODP0206-1256D-028-002/31-35
202.71
51.45
1.38
13.40
12.07
0.22
6.74
11.17
2.22
0.09
0.11
1425
516
99.26
ODP0206-1256D-029-001/98-99
207.58
51.77
1.41
13.43
12.30
0.21
6.85
11.17
2.32
0.09
0.12
1425
516
100.07
ODP0206-1256D-029-002/18-19
208.25
52.12
1.09
14.09
10.54
0.17
7.51
11.95
2.02
0.13
0.11
1233
696
100.12
ODP0206-1256D-030-001/3-4
211.43
52.24
1.02
13.96
10.58
0.19
7.59
11.90
1.96
0.09
0.09
1257
593
99.98
ODP0206-1256D-030-001/47-51
211.87
51.89
1.01
13.90
10.54
0.20
7.68
11.91
1.91
0.10
0.09
1261
621
99.63
ODP0206-1256D-030-001/70-71
212.10
52.13
1.01
13.94
10.62
0.19
7.71
11.94
2.01
0.09
0.09
1169
634
100.10
ODP0206-1256D-031-001/33-37
216.33
52.23
1.02
14.02
10.58
0.18
7.75
11.99
1.94
0.09
0.08
1236
619
100.25
ODP0206-1256D-031-001/50-60
216.50
51.59
1.42
13.61
12.23
0.23
6.67
11.23
2.23
0.09
0.11
1485
567
99.81
ODP0206-1256D-038-001/122-123
255.52
52.06
1.06
14.03
10.60
0.19
7.73
11.92
2.02
0.09
0.09
1139
479
100.12
ODP0206-1256D-038-002/61-63
256.41
51.93
1.07
14.10
10.50
0.19
7.59
11.94
1.97
0.09
0.09
1157
453
99.80
ODP0206-1256D-040-001/30-37
267.80
52.08
1.05
14.05
10.55
0.19
7.59
12.02
2.01
0.10
0.10
1187
490
100.09
ODP0206-1256D-040-001/86-102
268.36
51.07
1.06
13.98
10.62
0.19
7.71
11.95
2.09
0.09
0.08
1196
476
99.19
ODP0206-1256D-043-001/53-56
284.43
51.69
1.04
13.92
10.54
0.19
7.82
11.90
2.00
0.09
0.10
1226
748
99.68
ODP0206-1256D-043-001/7-12
283.97
51.31
1.05
14.00
10.60
0.19
7.75
11.96
2.00
0.10
0.09
1196
737
99.43
ODP0206-1256D-043-002/45-47
285.85
50.29
1.05
14.06
10.55
0.18
7.87
11.93
2.08
0.10
0.08
1187
709
98.56
ODP0206-1256D-044-001/124-125
294.34
51.05
1.08
13.93
10.62
0.20
7.82
11.89
2.04
0.09
0.09
1140
682
99.16
ODP0206-1256D-046-001/31-35
311.81
49.69
1.90
14.36
10.86
0.20
7.07
11.58
2.67
0.21
0.20
1384
256
99.11
ODP0206-1256D-047-001/116-117
322.16
50.23
1.34
14.16
10.54
0.20
7.80
12.31
2.27
0.09
0.09
1303
567
99.41
ODP0206-1256D-048-001/48-49
327.48
49.27
1.25
14.37
10.08
0.21
8.01
12.62
2.32
0.09
0.10
1252
456
98.67
ODP0206-1256D-048-001/112-113
328.12
50.18
1.26
14.38
10.12
0.20
8.08
12.58
2.32
0.09
0.11
1228
476
99.67
ODP0206-1256D-051-001/91-92
347.01
51.03
1.58
13.41
12.52
0.23
6.56
10.84
2.47
0.13
0.12
1490
643
99.31
ODP0206-1256D-051-001/135-136
347.45
50.84
1.58
13.45
12.60
0.23
6.61
10.84
2.49
0.12
0.12
1473
682
99.31
ODP0206-1256D-051-002/20-24
347.80
51.37
1.58
13.43
12.59
0.23
6.57
10.88
2.46
0.12
0.13
1570
673
99.82
ODP0206-1256D-051-002/131-132
348.08
51.29
1.58
13.49
12.52
0.25
6.62
10.86
2.44
0.12
0.14
1440
655
99.74
ODP0206-1256D-053-001/98-121
348.91
51.15
1.45
13.59
11.97
0.22
7.13
11.46
2.27
0.09
0.10
1474
445
99.84
ODP0206-1256D-053-002/68-69
360.78
51.44
1.43
13.62
12.07
0.22
7.24
11.51
2.26
0.09
0.10
1449
467
100.39
ODP0206-1256D-065-001/1-2
361.95
51.80
1.05
13.82
10.87
0.18
7.54
11.61
2.09
0.10
0.10
1203
658
99.53
ODP0206-1256D-070-001/61-65
455.71
51.20
1.27
13.80
11.85
0.21
7.14
11.50
2.24
0.08
0.09
1428
281
99.74
DiB
Depth in basement [m]
Table A4: Results of electron microprobe standard analyses.
sample ID
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
[wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %]
K 2O
P2O5
S
Cl
Total
[wt. %]
[wt. %]
[ppm]
[ppm]
[wt. %]
VG-2 preferred value
CAU Kiel (C. Beier, pers. comm.)
50.17
1.83
14.03
11.55
0.21
6.94
11.17
2.55
0.17
0.30
1452
262
99.08
Max Planck Institute for Chemistrya
50.70
1.87
14.12
11.80
0.21
6.60
10.92
2.84
0.20
0.22
1400
330
99.65
GEOMAR, Kiela
50.44
1.81
14.04
11.86
0.21
6.91
11.11
2.57
0.19
0.15
1400
400
99.47
GZN, FAU Erlangen-Nürnberg (this study)
50.27
1.85
14.01
11.84
0.16
6.97
11.09
2.62
0.18
0.19
1214
254
99.33
Smithsonian (1993): Regelous et al., 1999
50.57
1.85
14.06
11.59
0.22
7.07
11.12
2.63
0.19
0.20
Texas A&M Universitya
50.37
1.88
14.00
11.80
0.21
6.82
10.80
2.63
0.16
0.20
U.S. Geological Surveya
50.20
1.85
14.00
11.80
0.22
6.70
11.20
2.63
0.19
0.20
1200
270
99.14
Université Blaise Pascala
50.81
1.91
14.00
11.64
0.21
6.78
11.15
2.66
0.20
0.22
1571
296
99.77
University of Hawaii at Manoaa
50.66
1.86
14.04
11.86
0.21
6.71
11.15
2.66
0.19
0.22
1500
Utrecht Universitya
50.63
1.92
13.97
11.88
0.22
6.72
11.12
2.61
0.19
0.19
1416
303
99.62
mean
50.48
1.86
14.03
11.76
0.21
6.82
11.08
2.64
0.19
0.21
1394
302
99.41
99.50
98.87
99.71
VG-2 (GZN, FAU Erlangen-Nürnberg)
mean (n=90)
50.27
1.85
14.01
11.84
0.16
6.97
11.09
2.62
0.18
0.19
1214
254
99.33
Accuracy
-0.21
-0.01
-0.02
0.08
-0.05
0.15
0.01
-0.02
-0.01
-0.02
-180
-48
-0.09
Rel. accuracy (%)b
-0.42
-0.69
-0.12
0.66
-23.18
2.17
0.07
-0.75
-3.16
-8.95
-12.92
-15.94
-0.09
Std. dev. (1σ; n=90)
0.66
0.05
0.25
0.35
0.09
0.15
0.15
0.11
0.02
0.07
385
33
0.80
Std. err.c
0.09
0.02
0.05
0.06
0.03
0.04
0.04
0.04
0.01
0.03
2.07
0.61
0.09
Smithsonian (2002): Melson et al., 2002
50.81
1.85
14.06
11.84
0.22
6.71
11.12
2.62
0.19
0.20
1415d
301d
Correction to VG-2 preferred value
0.9935 1.0070 0.9976 0.9934 0.9467 1.0167 0.9966 1.0076 0.9782 1.0434 0.9853 1.0039
Correction factors (Interlab bias)
GZN, FAU Erlangen-Nürnberg (this study)
50.27
Correction to VG-2 preferred value
1.0042 1.0070 1.0012 0.9934 1.3017 0.9787 0.9993 1.0076 1.0326 1.0983 1.1484 1.1897
1.85
14.01
11.84
0.16
6.97
11.09
2.62
0.18
0.19
1214
254
99.79
0.9962
99.33
1.0009
VG-A99 (GZN, FAU Erlangen-Nürnberg)e
VG-A99: Jarosewich et al., 1980
50.94
4.06
12.49
13.30
0.15
5.08
9.30
2.66
0.82
0.42
135f
227f
99.26
mean (n=164)e
51.20
4.11
12.49
13.35
0.18
4.96
9.21
2.71
0.85
0.44
145
240
99.42
Accuracy
0.26
0.05
0.00
0.05
0.03
-0.12
-0.09
0.05
0.03
0.02
10
13
0.16
Rel. accuracy (%)b
0.50
1.16
0.03
0.37
22.93
-2.37
-0.92
1.72
3.34
4.01
7.56
5.90
0.16
Std. dev. (1σ; n=164)
0.53
0.08
0.24
0.36
0.09
0.13
0.13
0.11
0.07
0.14
84
39
0.65
Std. err.c
0.06
0.02
0.04
0.05
0.02
0.03
0.03
0.03
0.02
0.03
0.72
0.49
0.06
a
data source: GeoReM, http://georem.mpch-mainz.gwdg.de
b
Relative accuracy: (Measured standard - accepted standard) / (accepted standard) in % as calculated by Fisk and Kelley (2002)
c
Standard error: √ (Std. dev. / n) as calculated by Jenner and O'Neill (2012)
d
data reported by Jenner and O'Neill (2012)
e
results of VG-A99 analyses (n=164) have been corrected for Interlab bias
f
data reported by Thordarsson et al. (1996)
Table A5: Averages for 500 m ridge depth intervals of the global mid-ocean ridge system.
Interval
RMWDa
K/Ti
1σ
Mg#b
1σ
[m]
Ti8.0
1σ
[wt. %]
Fe8.0
1σ
[wt. %]
Na8.0
1σ
CaO/Al2O3
1σ
nc
[wt. %]
500-1000 m below sealevel
-736
0.115 0.041 0.533 0.039
1.13
0.14
10.79
0.38
1.81
1.59
0.8301
0.0303
171
1000-1500 m below sealevel
-1306
0.111 0.076 0.572 0.073
1.09
0.19
10.22
1.21
2.13
0.45
0.8183
0.0558
120
1500-2000 m below sealevel
-1813
0.152 0.087 0.593 0.058
1.28
0.24
9.84
0.91
2.45
0.36
0.7900
0.3043
363
2000-2500 m below sealevel
-2322
0.136 0.071 0.590 0.069
1.27
0.18
9.88
0.99
2.41
0.28
0.7920
0.0450
954
2500-3000 m below sealevel
-2678
0.120 0.058 0.580 0.067
1.39
0.19
9.74
0.74
2.54
0.26
0.7886
0.0387 3258
3000-3500 m below sealevel
-3201
0.126 0.071 0.594 0.057
1.41
0.20
9.67
0.85
2.62
0.29
0.7547
0.0533 1062
3500-4000 m below sealevel
-3768
0.141 0.088 0.621 0.035
1.44
0.19
9.15
1.04
2.88
0.29
0.7140
0.0523 1004
4000-4500 m below sealevel
-4151
0.134 0.066 0.621 0.034
1.45
0.18
8.88
0.92
2.95
0.38
0.7103
0.0493
364
4500-5000 m below sealevel
-4777
0.161 0.062 0.610 0.044
1.49
0.19
8.86
1.00
3.13
0.44
0.6822
0.0560
184
5000-5500 m below sealevel
-5153
0.179 0.042 0.601 0.039
1.41
0.16
7.86
0.92
3.36
0.42
0.6605
0.0575
82
a
real mean water depth (RMWD) of the interval
b
Calculated assuming an Fe2+/∑Fe of 0.9
c
Number of samples with MgO > 7.0 wt. %
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
DSDP011-0105-041-002/117-119
Atlantic
156
1.09
10.12
2.00
0.8481
DSDP011-0105-041-003/46-48
Atlantic
156
1.24
11.24
2.25
0.8358
DSDP011-0105-041-003/108-110
Atlantic
156
1.18
11.06
2.22
0.8409
DSDP011-0105-042-001/50-52
Atlantic
156
1.18
10.99
2.24
0.8372
DSDP011-0105-042-001/122-124
Atlantic
156
1.23
11.20
2.23
0.8475
DSDP011-0105-042-002//88-89
Atlantic
156
1.22
11.12
2.19
0.8364
DSDP011-0105-043-001/77-79
Atlantic
156
1.22
11.12
2.24
0.8441
DSDP011-0105-043-002/25-26
Atlantic
156
1.22
11.15
2.26
0.8407
DSDP045-0396-014-006/14-15
Atlantic
13
1.41
8.29
2.35
0.7189
DSDP045-0396-015-001/44-46
Atlantic
13
1.40
8.90
2.67
0.7154
DSDP045-0396-015-001/136-137
Atlantic
13
1.44
8.52
2.46
0.7286
DSDP045-0396-015-003/28-30
Atlantic
13
1.40
8.90
2.64
0.7178
DSDP045-0396-015-004/100-102
Atlantic
13
1.37
8.80
2.57
0.7107
DSDP045-0396-016-002/80-83
Atlantic
13
1.40
8.83
2.70
0.7095
DSDP045-0396-016-004/56-57
Atlantic
13
1.41
8.91
2.72
0.7084
DSDP045-0396-018-001/38-40
Atlantic
13
1.36
8.70
2.54
0.7191
DSDP045-0396-018-CC/25-27
Atlantic
13
1.42
8.99
2.73
0.7067
DSDP045-0396-019-002/61-63
Atlantic
13
1.47
9.11
2.36
0.7012
DSDP045-0396-021-001/93-95
Atlantic
13
1.35
8.38
2.45
0.7105
DSDP045-0396-022-001/115-117
Atlantic
13
1.42
8.91
2.72
0.7134
DSDP045-0396-022-003/5-6
Atlantic
13
1.41
8.98
2.67
0.7135
DSDP045-0396-022-004/98-100
Atlantic
13
1.35
8.82
2.35
0.7567
DSDP045-0396-024-003/72-76
Atlantic
13
1.40
8.91
2.40
0.7530
DSDP051-0417D-022-001/95-96
Atlantic
120
1.43
10.27
2.19
0.8225
DSDP051-0417D-022-007/28-30
Atlantic
120
1.44
10.38
2.32
0.8083
DSDP051-0417D-026-001/50-54
Atlantic
120
1.43
10.30
2.32
0.8147
DSDP051-0417D-026-002/42-44
Atlantic
120
1.46
10.39
2.27
0.8205
DSDP051-0417D-026-006/47-51
Atlantic
120
1.41
10.32
2.33
0.8018
DSDP051-0417D-026-007/3-5
Atlantic
120
1.40
10.20
2.35
0.8016
DSDP051-0417D-027-002/5-7
Atlantic
120
1.44
10.36
2.35
0.7995
DSDP051-0417D-027-003/27-29
Atlantic
120
1.44
10.40
2.31
0.7967
DSDP051-0417D-027-004/51-53
Atlantic
120
1.44
10.43
2.37
0.7977
DSDP051-0417D-027-005/56-59
Atlantic
120
1.42
10.30
2.38
0.8003
DSDP051-0417D-027-006/123-125
Atlantic
120
1.46
10.47
2.37
0.7981
DSDP051-0417D-027-007/3-7
Atlantic
120
1.47
10.35
2.32
0.8017
DSDP051-0417D-028-001/104-105
Atlantic
120
1.43
10.35
2.37
0.8041
DSDP051-0417D-028-002/16-18
Atlantic
120
1.44
10.35
2.35
0.8010
DSDP051-0417D-028-002/85-87
Atlantic
120
1.44
10.33
2.40
0.8019
DSDP051-0417D-028-003/53-57
Atlantic
120
1.44
10.43
2.37
0.7968
DSDP051-0417D-028-004/63-64
Atlantic
120
1.43
10.33
2.34
0.8045
DSDP051-0417D-028-006/8-10
Atlantic
120
1.42
10.39
2.39
0.7984
DSDP051-0417D-029-001/108-110
Atlantic
120
1.44
10.31
2.37
0.8033
DSDP051-0417D-029-002/126-129
Atlantic
120
1.41
10.27
2.34
0.7955
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
DSDP051-0417D-029-003/120-122
Atlantic
120
1.45
10.47
2.42
0.8011
DSDP051-0417D-029-004/54-56
Atlantic
120
1.42
10.40
2.39
0.7975
DSDP051-0417D-029-006/95-100
Atlantic
120
1.44
10.38
2.35
0.7999
DSDP051-0417D-030-001/88-91
Atlantic
120
1.40
10.26
2.30
0.8037
DSDP051-0417D-030-002/99-100
Atlantic
120
1.44
10.35
2.35
0.8065
DSDP051-0417D-030-003/129-132
Atlantic
120
1.47
10.49
2.39
0.8059
DSDP051-0417D-030-004/54-56
Atlantic
120
1.43
10.35
2.36
0.8029
DSDP051-0417D-030-005/94-96
Atlantic
120
1.43
10.47
2.38
0.8051
DSDP051-0417D-030-006/36-38
Atlantic
120
1.42
10.33
2.35
0.8066
DSDP051-0417D-031-001/11-15
Atlantic
120
1.42
10.27
2.28
0.8089
DSDP051-0417D-031-002/129-131
Atlantic
120
1.44
10.36
2.32
0.8061
DSDP051-0417D-031-004/17-18
Atlantic
120
1.46
10.36
2.35
0.8023
DSDP051-0417D-034-005/118-120
Atlantic
120
1.42
10.12
2.23
0.8225
DSDP051-0417D-035-005/126-128
Atlantic
120
1.32
9.88
2.08
0.8019
DSDP051-0417D-037-001/34-36
Atlantic
120
1.33
10.01
2.11
0.8097
DSDP051-0417D-037-004/33-36
Atlantic
120
1.31
9.87
2.08
0.7983
DSDP051-0417D-037-007/25-28
Atlantic
120
1.36
10.08
2.13
0.8065
DSDP051-0417D-038-002/4-6
Atlantic
120
1.36
10.09
2.10
0.8041
DSDP051-0417D-039-006/57-58
Atlantic
120
1.30
9.93
2.10
0.8200
DSDP051-0417D-040-003/7-9
Atlantic
120
1.31
9.88
2.11
0.8224
DSDP051-0417D-041-004/107-109
Atlantic
120
1.21
9.12
2.06
0.8135
DSDP051-0417D-042-002/8-12
Atlantic
120
1.24
9.75
2.08
0.8126
DSDP052-0417D-060-006/39-40
Atlantic
120
1.31
9.88
2.17
0.7909
DSDP052-0417D-062-003/12-13
Atlantic
120
1.33
9.82
2.08
0.7872
DSDP052-0417D-062-006/81-83
Atlantic
120
1.34
10.04
2.13
0.7844
DSDP052-0417D-063-004/133-135
Atlantic
120
1.34
10.03
2.12
0.7833
DSDP052-0417D-063-006/79-80
Atlantic
120
1.32
9.84
2.11
0.7778
DSDP052-0417D-064-002/128-130
Atlantic
120
1.37
10.10
2.18
0.7891
DSDP052-0417D-064-004/40-43
Atlantic
120
1.34
10.08
2.16
0.7944
DSDP052-0417D-066-002/28-30
Atlantic
120
1.32
10.00
2.16
0.7771
DSDP052-0417D-066-004/80-83
Atlantic
120
1.32
9.92
2.17
0.7871
DSDP052-0417D-066-005/33-34
Atlantic
120
1.33
9.94
2.21
0.7879
DSDP052-0417D-066-006/41-43
Atlantic
120
1.25
9.18
2.04
0.7678
DSDP052-0418A-015-001/64-66
Atlantic
120
1.21
9.78
2.18
0.8272
DSDP052-0418A-015-001/140-142
Atlantic
120
1.19
9.80
2.10
0.8380
DSDP052-0418A-043-001/15-17
Atlantic
120
1.21
10.19
2.21
0.8014
DSDP052-0418A-045-001/100-102
Atlantic
120
1.25
10.21
2.14
0.8077
DSDP052-0418A-048-001/56-58
Atlantic
120
1.27
10.32
2.24
0.7967
DSDP053-0418A-055-001/5-6
Atlantic
120
1.29
9.89
1.87
0.8373
DSDP053-0418A-059-004/3-4
Atlantic
120
1.37
9.89
1.94
0.7948
DSDP053-0418A-064-003/53-55
Atlantic
120
1.28
9.83
2.07
0.8350
DSDP053-0418A-068-001/133-134
Atlantic
120
1.35
10.10
2.01
0.8267
DSDP053-0418A-072-001/50-52
Atlantic
120
1.32
9.97
1.97
0.8312
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
DSDP053-0418A-074-004/44-45
Atlantic
120
1.27
9.63
1.86
0.8216
DSDP053-0418A-075-005/3-6
Atlantic
120
1.26
9.88
2.04
0.8177
DSDP053-0418A-086-004/32-34
Atlantic
120
1.28
9.87
1.92
0.8195
DSDP073-0522B-003-002/117-119
Atlantic
37.1
1.42
8.77
2.62
0.8533
DSDP073-0522B-003-002/128-134
Atlantic
37.1
1.43
8.65
2.56
0.8540
DSDP073-0522B-003-003/75-78
Atlantic
37.1
1.42
8.67
2.60
0.8462
DSDP073-0522B-003-003/117-121
Atlantic
37.1
1.38
8.30
2.51
0.8459
DSDP073-0522B-003-004/4-7
Atlantic
37.1
1.45
8.87
2.09
0.8414
DSDP073-0522B-004-001/82-85
Atlantic
37.1
1.28
7.10
2.94
0.8247
DSDP073-0522B-004-002/56-57
Atlantic
37.1
1.42
8.79
2.26
0.8504
DSDP073-0522B-005-001/11-13
Atlantic
37.1
1.38
8.69
2.56
0.8476
DSDP073-0522B-005-001/113-115
Atlantic
37.1
1.38
8.66
2.61
0.8319
DSDP073-0522B-005-002/6-7
Atlantic
37.1
1.40
8.66
2.53
0.8334
DSDP073-0522B-005-003/139-141
Atlantic
37.1
1.39
8.59
2.52
0.8482
DSDP073-0522B-006-001/10-12
Atlantic
37.1
1.28
9.33
2.40
0.8729
DSDP076-0534A-128-004/136-139
Atlantic
162
0.78
9.55
1.58
0.8540
DSDP076-0534A-129-005/6-8
Atlantic
162
0.84
10.02
1.69
0.8524
DSDP076-0534A-130-001/72-74
Atlantic
162
0.86
10.02
1.70
0.8502
DSDP076-0534A-130-001/100-101
Atlantic
162
0.84
9.93
1.67
0.8533
DSDP078-0543A-013-001/10-11
Atlantic
80
1.42
8.72
2.50
0.7928
DSDP078-0543A-013-002/2-3
Atlantic
80
1.38
8.59
2.47
0.7844
DSDP078-0543A-013-002/105-106
Atlantic
80
1.43
8.67
2.50
0.7767
DSDP078-0543A-013-004/125-126
Atlantic
80
1.40
8.71
2.46
0.7805
DSDP078-0543A-013-005/98-99
Atlantic
80
1.41
8.76
2.54
0.7971
DSDP078-0543A-014-001/57-59
Atlantic
80
1.39
8.68
2.55
0.7831
DSDP078-0543A-015-002/2-3
Atlantic
80
1.40
8.56
2.53
0.7743
DSDP078-0543A-015-003/137-139
Atlantic
80
1.37
8.62
2.48
0.7959
DSDP078-0543A-015-005/18-21
Atlantic
80
1.40
8.77
2.55
0.7961
DSDP078-0543A-016-001/139-140
Atlantic
80
1.44
8.77
2.57
0.7990
DSDP078-0543A-016-003/57-59
Atlantic
80
1.46
8.57
2.55
0.7725
DSDP078-0543A-016-004/116-117
Atlantic
80
1.41
8.45
2.50
0.7664
DSDP078-0543A-016-006/2-3
Atlantic
80
1.49
8.69
2.62
0.7659
DSDP078-0543A-016-006/118-120
Atlantic
80
1.49
8.85
2.57
0.7746
DSDP078-0543A-016-007/105-106
Atlantic
80
1.44
8.64
2.56
0.7741
DSDP016-0163-028-001/113-114
Pacific
72
1.24
9.65
2.53
0.7857
DSDP016-0163-029-001/90-92
Pacific
72
1.34
10.54
2.62
0.7976
DSDP029-0278-035-001/78-80
Pacific
30
1.17
9.37
2.55
0.8421
DSDP029-0278-035-001/111-112
Pacific
30
1.19
9.54
2.62
0.8396
DSDP029-0278-035-002/90-92
Pacific
30
1.18
9.40
2.52
0.8362
DSDP029-0278-035-003/10-12
Pacific
30
1.19
9.71
2.61
0.8356
DSDP029-0278-035-003/68-69
Pacific
30
1.19
9.63
2.64
0.8340
DSDP063-0469-048-001/54-57
Pacific
17
1.42
10.17
2.17
0.8338
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
DSDP063-0469-049-001/106-107
Pacific
17
1.49
10.38
2.28
0.8308
DSDP063-0469-050-001/80-81
Pacific
17
1.45
10.19
2.28
0.7995
DSDP063-0469-050-001/111-112
Pacific
17
1.47
10.37
2.31
0.7950
DSDP063-0469-050-002/11-12
Pacific
17
1.47
10.43
2.35
0.8058
DSDP063-0470A-007-001/67-68
Pacific
15.5
1.51
9.74
2.25
0.8185
DSDP063-0470A-007-002/23-24
Pacific
15.5
1.38
9.50
2.43
0.7589
DSDP063-0470A-007-003/53-54
Pacific
15.5
1.40
9.37
2.46
0.7595
DSDP063-0470A-008-001/59-60
Pacific
15.5
1.41
9.54
2.72
0.7884
DSDP063-0470A-008-002/26-27
Pacific
15.5
1.39
9.45
2.72
0.7865
DSDP063-0470A-008-003/112-113
Pacific
15.5
1.37
9.37
2.66
0.7812
DSDP063-0470A-008-004/142-143
Pacific
15.5
1.42
9.60
2.79
0.7865
DSDP063-0470A-008-005/63-65
Pacific
15.5
1.43
9.66
2.76
0.7843
DSDP063-0470A-009-001/81-82
Pacific
15.5
1.41
9.51
2.69
0.7818
DSDP063-0470A-011-001/3-4
Pacific
15.5
1.40
9.39
2.67
0.7890
DSDP063-0470A-011-001/64-65
Pacific
15.5
1.30
9.55
2.39
0.8057
DSDP063-0470A-012-001/34-35
Pacific
15.5
1.68
8.85
2.55
0.7156
DSDP063-0470A-013-001/40-41
Pacific
15.5
1.57
9.76
2.27
0.8059
DSDP063-0472-014-001/13-14
Pacific
15
1.30
9.01
2.68
0.7830
DSDP068-0501-010-001/49-50
Pacific
5.9
1.05
9.74
2.07
0.8771
DSDP068-0501-014-003/138-140
Pacific
5.9
1.13
10.31
2.28
0.8563
DSDP068-0501-014-004/61-63
Pacific
5.9
1.09
10.26
2.25
0.8545
DSDP068-0501-015-001/76-77
Pacific
5.9
1.08
10.29
2.26
0.8538
DSDP068-0501-015-003/64-65
Pacific
5.9
1.12
10.36
2.28
0.8519
DSDP068-0501-015-004/2-3
Pacific
5.9
1.10
10.16
2.29
0.8545
DSDP068-0501-017-002/113-115
Pacific
5.9
1.06
9.90
2.16
0.8498
DSDP068-0501-020-003/121-122
Pacific
5.9
1.00
9.99
2.02
0.8834
DSDP069-0504B-004-001/66-67
Pacific
5.9
1.12
10.40
2.22
0.8301
DSDP069-0504B-005-002/78-79
Pacific
5.9
1.09
10.17
2.19
0.8522
DSDP069-0504B-006-002/142-144
Pacific
5.9
1.14
10.39
2.26
0.8462
DSDP069-0504B-012-002/38-40
Pacific
5.9
1.10
9.98
2.38
0.8484
DSDP069-0504B-015-002/51-52
Pacific
5.9
1.07
10.55
2.13
0.8632
DSDP069-0504B-015-005/64-66
Pacific
5.9
1.04
10.48
2.09
0.8595
DSDP069-0504B-016-003/123-124
Pacific
5.9
1.12
9.57
2.14
0.8581
DSDP069-0504B-018-001/10-12
Pacific
5.9
1.20
10.19
2.44
0.8517
DSDP069-0504B-021-004/108-109
Pacific
5.9
1.01
10.42
2.14
0.8907
DSDP069-0504B-022-002/29-30
Pacific
5.9
0.98
10.30
2.11
0.8816
DSDP069-0504B-025-001/102-104
Pacific
5.9
1.10
10.04
2.27
0.8828
DSDP069-0504B-026-001/93-94
Pacific
5.9
1.10
9.95
2.18
0.8836
DSDP069-0504B-028-003/42-43
Pacific
5.9
1.14
10.34
2.20
0.8755
DSDP069-0504B-029-002/111-112
Pacific
5.9
1.11
10.03
2.28
0.8904
DSDP070-0504B-035-001/17-19
Pacific
5.9
1.13
11.06
2.81
0.7650
DSDP070-0504B-038-001/6-7
Pacific
5.9
1.04
10.33
2.25
0.8798
DSDP070-0504B-045-002/23-24
Pacific
5.9
1.10
10.33
2.03
0.8607
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
DSDP070-0504B-048-002/123-124
Pacific
5.9
1.03
10.08
2.00
0.8821
DSDP070-0504B-049-003/47-48
Pacific
5.9
1.02
10.19
2.00
0.8876
DSDP070-0504B-056-001/87-88
Pacific
5.9
1.41
9.69
2.50
0.8300
DSDP070-0504B-061-001/31-32
Pacific
5.9
1.20
10.69
2.08
0.8511
DSDP070-0504B-064-001/29-30
Pacific
5.9
1.14
10.18
2.14
0.8769
DSDP085-0573B-043-001/27-28
Pacific
35.5
1.59
9.83
2.57
0.8162
DSDP091-0595B-007-002/67-70
Pacific
90
1.44
9.89
2.48
0.8200
DSDP091-0595B-007-002/80-81
Pacific
90
1.48
9.88
2.50
0.8112
DSDP092-0597A-007-CC/5-6
Pacific
28.3
1.02
10.55
2.21
0.8789
ODP0185-0801C-027-003/2-3
Pacific
166
1.52
10.16
2.27
0.8185
ODP0185-0801C-042-002/73-89
Pacific
166
1.41
10.28
2.50
0.7873
ODP0185-0801C-042-002/90-94
Pacific
166
1.40
10.22
2.61
0.8017
ODP0185-0801C-042-002/94-100
Pacific
166
1.43
10.44
2.61
0.8134
ODP0185-1149D-009-002/71-72
Pacific
132
1.45
10.06
2.09
0.8137
ODP0191-1179D-019-003/126-127
Pacific
129
1.33
10.48
2.17
0.8233
ODP0191-1179D-020-002/97-99
Pacific
129
1.65
10.60
2.71
0.7381
ODP0191-1179D-020-005/56-60
Pacific
129
1.71
10.77
2.73
0.7272
ODP0191-1179D-021-001/94-95
Pacific
129
1.71
10.78
2.74
0.7191
ODP0191-1179D-021-002/52-53
Pacific
129
1.74
10.96
2.76
0.7280
ODP0191-1179D-021-003/38-41
Pacific
129
1.77
10.94
2.85
0.7230
ODP0191-1179D-021-004/64-67
Pacific
129
1.71
10.60
2.75
0.7175
ODP0191-1179D-022-001/104-105
Pacific
129
1.67
10.25
2.68
0.7086
ODP0191-1179D-022-001/91-92
Pacific
129
1.72
10.49
2.69
0.7162
ODP0191-1179D-022-003/46-51
Pacific
129
1.64
10.67
2.65
0.7519
ODP0191-1179D-022-003/55-56
Pacific
129
1.81
11.15
2.88
0.7177
ODP0199-1217A-017-CC/1-8
Pacific
48
1.16
9.28
2.41
0.8510
ODP0199-1217A-017-CC/9-13
Pacific
48
1.17
9.25
2.37
0.8562
ODP0203-1243B-006-002/1-6
Pacific
10
1.40
8.45
2.87
0.8012
ODP0203-1243B-011-001/12-21
Pacific
10
1.50
8.56
2.64
0.7571
ODP0206-1256D-015-002/27-28
Pacific
15
0.92
9.98
1.84
0.8543
ODP0206-1256D-015-003/2-3
Pacific
15
0.87
9.52
1.71
0.8481
ODP0206-1256D-024-001/83-92
Pacific
15
0.97
9.61
1.83
0.8551
ODP0206-1256D-024-002/20-25
Pacific
15
0.97
9.58
1.81
0.8569
ODP0206-1256D-024-002/102-106
Pacific
15
0.96
9.77
1.83
0.8601
ODP0206-1256D-029-002/18-19
Pacific
15
0.95
9.45
1.81
0.8387
ODP0206-1256D-030-001/3-4
Pacific
15
0.90
9.63
1.77
0.8434
ODP0206-1256D-030-001/47-51
Pacific
15
0.92
9.74
1.76
0.8477
ODP0206-1256D-030-001/70-71
Pacific
15
0.92
9.86
1.87
0.8469
ODP0206-1256D-031-001/33-37
Pacific
15
0.94
9.89
1.81
0.8461
ODP0206-1256D-038-001/122-123
Pacific
15
0.97
9.88
1.88
0.8401
ODP0206-1256D-038-002/61-63
Pacific
15
0.95
9.55
1.78
0.8381
ODP0206-1256D-040-001/30-37
Pacific
15
0.93
9.60
1.82
0.8465
ODP0206-1256D-040-001/86-102
Pacific
15
0.97
9.86
1.94
0.8459
Table A6: Fractionation corrected data (ancient MORB) with MgO ≥7.0 wt. % normalised to MgO = 8.0 wt. %.
sample ID
Ocean
Age
[Ma]
Ti8
Fe8
Na8
CaO/Al2O3
[wt. %] [wt. %] [wt. %]
ODP0206-1256D-043-001/7-12
Pacific
15
0.97
9.90
1.87
0.8455
ODP0206-1256D-043-001/53-56
Pacific
15
0.97
9.96
1.90
0.8461
ODP0206-1256D-043-002/45-47
Pacific
15
0.99
10.05
2.00
0.8397
ODP0206-1256D-044-001/124-125
Pacific
15
1.01
10.04
1.94
0.8439
ODP0206-1256D-047-001/116-117
Pacific
15
1.24
9.92
2.16
0.8599
ODP0206-1256D-048-001/48-49
Pacific
15
1.22
9.80
2.29
0.8688
ODP0206-1256D-048-001/112-113
Pacific
15
1.24
9.96
2.31
0.8658
ODP0206-1256D-053-002/68-69
Pacific
15
1.17
10.55
1.95
0.8361
ODP0206-1256D-065-001/1-2
Pacific
15
0.92
9.84
1.89
0.8313
Table A7: Major element, trace element and Sr-Nd-Pb isotope analyses of Seamount 6 lavas.
sample
3009-1511
3009-1557
3010-1032
3010-1053
3010-1219
3010-1249
3010-1453
3010-1502
3011-1333
Tr
Th
Th
Th
Th
AB
Th
AB
Th
12.725
102.584
-1892
12.725
102.585
-1837
12.720
102.585
-1977
12.721
102.584
-1976
12.726
102.584
-1884
12.729
102.584
-1842
12.738
102.585
-1695
12.738
102.585
-1690
12.720
102.584
-1975
(wt. %)
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K2 O
P2O5
SO3
F
Cl
Total
50.00
1.35
16.13
8.80
0.15
7.92
11.93
2.88
0.22
0.15
0.27
0.08
0.03
99.87
48.91
0.92
17.22
8.17
0.14
9.72
12.63
2.43
0.02
0.56
0.02
0.01
0.00
100.74
49.98
1.19
16.30
8.66
0.17
8.62
12.16
2.77
0.10
0.10
0.26
0.10
0.01
100.37
50.07
1.17
16.25
8.74
0.17
8.70
12.04
2.78
0.10
0.11
0.27
0.07
0.00
100.45
49.76
1.08
16.70
8.47
0.16
8.83
12.22
2.70
0.07
0.10
0.25
0.18
0.01
100.44
50.58
2.35
17.00
8.27
0.15
5.03
8.77
4.57
1.80
0.70
0.23
0.08
0.06
99.56
50.06
1.18
16.27
8.64
0.16
8.60
12.12
2.77
0.10
0.09
0.27
0.05
0.01
100.29
50.41
2.33
17.87
7.83
0.14
4.76
8.51
4.73
2.06
0.56
0.02
0.06
0.08
99.30
49.66
1.16
16.36
8.61
0.16
8.70
12.02
2.78
0.09
0.11
0.26
0.07
0.00
99.98
Mg#
Si72
Na72
Fe72
0.64
48.98
2.78
7.82
0.70
48.80
2.46
8.22
0.66
50.54
2.84
9.11
0.66
49.31
2.71
8.16
0.67
49.21
2.63
8.03
0.55
50.31
4.16
4.76
0.66
49.31
2.70
8.05
0.55
50.14
4.31
4.31
0.67
50.14
2.85
9.01
(ppm)
Li
Sc
V
Cr
Co
Ni
Cu
Zn
Ga
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Tl
Pb
Th
U
5.67
38.6
242
216
43.4
88.6
87.9
66.7
16.7
3.75
169
28.8
99.4
6.80
0.53
0.82
0.02
0.04
43.6
5.60
14.0
2.10
10.5
3.26
1.23
4.06
0.73
4.75
1.00
2.75
0.41
2.66
0.39
2.39
0.41
0.15
0.03
0.57
0.49
0.16
5.50
38.4
239
205
44.8
103
90.1
64.3
16.0
1.21
134
27.1
76.2
2.57
0.30
0.68
0.02
0.02
13.5
2.87
8.47
1.45
7.95
2.79
1.09
3.64
0.68
4.46
0.94
2.60
0.38
2.53
0.37
1.95
0.17
0.16
0.10
0.37
0.17
0.06
5.19
38.8
238
296
46.7
122
93.4
63.0
16.0
0.71
123
26.1
67.3
1.66
0.18
0.61
0.01
0.01
8.70
2.17
6.87
1.24
7.14
2.60
1.03
3.44
0.65
4.28
0.91
2.50
0.37
2.42
0.36
1.78
0.11
0.04
0.01
0.31
0.10
0.04
6.76
30.4
209
78.2
29.1
25.9
63.3
69.5
19.9
37.9
406
37.3
312
57.9
3.58
2.10
0.08
0.41
366
35.7
70.0
8.36
33.6
7.14
2.28
7.20
1.12
6.51
1.28
3.47
0.50
3.26
0.48
6.35
3.28
0.80
0.07
2.60
4.54
1.35
5.50
40.1
245
200
45.5
104
92.6
63.2
16.7
1.21
136
27.5
78.0
2.59
0.24
0.68
0.01
0.01
13.4
2.90
8.55
1.46
8.07
2.81
1.10
3.67
0.68
4.50
0.95
2.63
0.39
2.56
0.38
1.97
0.17
0.05
0.01
0.36
0.17
0.06
6.70
24.6
181
71.0
28.2
40.6
51.1
67.5
20.1
41.6
478
35.0
323
63.4
3.92
2.14
0.10
0.45
414
38.9
76.3
8.99
35.5
7.28
2.31
7.10
1.09
6.15
1.20
3.23
0.46
3.03
0.45
6.55
3.60
0.88
0.08
2.87
4.99
1.48
rock type
Lat. (°N)
Long. (°W)
WD (m)
0.702709±9
Sr/86Sr
Nd/144Nd 0.513076±8
εNd
8.4
206
Pb/204Pb 18.5296±31
207
Pb/204Pb 15.5309±38
208
Pb/204Pb 38.0757±124
87
143
0.702577±4 0.702585±23 0.702970±6 0.702570±6 0.703006±4
0.513160±18 0.513150±10 0.512958±4 0.513138±12 0.512947±11
10.1
9.9
6.2
9.7
6.0
18.3210±29 18.2408±22 18.9183±17
lead lost
18.9487±27
15.4943±36 15.4843±26 15.5951±20
lead lost
15.6004±32
37.8344±118 37.7505±85 38.5326±66
lead lost
38.5689±104
Errors for isotope data are 2SD and refer to the last digit.
Table A7: Major element, trace element and Sr-Nd-Pb isotope analyses of Seamount 6 lavas.
sample
3013-1315
3013-1425
3014-1136
3015-1048
3015-1344
3015-1432
3015-1437
3016-1502
VG-2
AB
AB
Th
Tr
AB
AB
AB
AB
n=49
12.720
102.584
-1976
12.720
102.584
-1975
12.724
102.590
-1913
12.725
102.578
-1930
12.735
102.575
-1847
12.736
102.580
-1776
12.735
102.580
-1776
12.738
102.587
-1746
(wt. %)
SiO2
TiO2
Al2O3
FeOT
MnO
MgO
CaO
Na2O
K2 O
P2O5
SO3
F
Cl
Total
51.19
2.48
15.68
10.00
0.19
4.83
8.82
4.64
1.45
0.56
0.02
0.05
0.06
99.95
50.00
2.33
17.86
7.73
0.15
4.98
8.33
4.71
2.00
0.73
0.21
0.11
0.07
99.16
50.34
1.18
16.32
8.58
0.16
8.68
12.03
2.79
0.10
0.10
0.26
0.07
0.01
100.59
50.10
1.88
15.69
9.23
0.18
6.82
11.05
3.57
0.57
0.25
0.27
0.07
0.01
99.68
48.22
1.81
18.61
7.73
0.15
7.37
9.95
3.54
1.03
0.40
0.19
0.14
0.03
99.11
48.26
1.81
18.68
7.69
0.13
7.52
9.89
3.52
1.01
0.40
0.19
0.13
0.02
99.21
48.15
1.79
18.59
7.62
0.12
7.60
9.91
3.54
1.01
0.41
0.19
0.12
0.02
99.01
49.62
2.23
17.26
8.03
0.15
5.31
8.85
4.52
1.68
0.65
0.24
0.09
0.06
98.62
Mg#
Si72
Na72
Fe72
0.49
50.61
4.03
4.20
0.56
49.65
4.35
4.70
0.67
50.81
2.85
8.98
0.59
49.36
3.35
7.19
0.65
47.31
3.46
6.99
0.66
47.43
3.45
7.05
0.66
47.40
3.46
7.04
0.57
49.21
4.20
5.19
(ppm)
Li
Sc
V
Cr
Co
Ni
Cu
Zn
Ga
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Tl
Pb
Th
U
7.73
31.4
270
12.2
32.7
11.6
50.2
86.2
20.3
29.3
321
43.0
292
46.1
2.90
2.08
0.07
0.33
287
29.4
59.9
7.43
31.0
7.11
2.33
7.53
1.23
7.42
1.50
4.09
0.59
3.90
0.58
6.16
2.65
0.63
0.06
2.22
3.60
1.07
6.38
22.9
169
80.2
25.8
37.2
46.6
61.4
38.4
40.5
468
31.6
304
55.3
3.66
6.19
43.4
275
242
39.3
50.7
83.4
73.3
18.0
6.96
215
36.1
165
15.1
1.10
1.25
0.03
0.08
70.8
10.7
25.2
3.50
16.3
4.58
1.63
5.41
0.96
6.10
1.26
3.48
0.51
3.34
0.50
3.62
0.89
0.17
0.03
0.98
0.98
0.33
4.86
22.6
160
124
34.4
134
48.9
57.2
25.7
17.6
358
25.3
180
27.0
1.87
1.43
2.15
0.68
5.32
24.1
173
140
37.3
144
52.1
65.3
17.9
17.7
377
28.2
193
30.9
1.94
1.38
0.04
0.19
200
19.6
41.1
5.15
21.7
4.97
1.71
5.20
0.83
4.91
0.98
2.63
0.38
2.47
0.36
4.05
1.77
0.39
0.04
1.45
2.21
0.68
5.72
24.5
177
104
38.1
145
52.0
62.7
18.2
18.0
387
28.5
196
31.6
1.98
1.39
0.05
0.20
202
19.7
41.6
5.16
21.8
4.98
1.70
5.17
0.83
4.89
0.97
2.62
0.37
2.44
0.36
4.04
1.79
0.39
0.04
1.48
2.20
0.68
6.47
29.6
206
119
30.3
31.9
62.2
67.8
19.8
34.9
414
35.8
297
54.5
3.34
1.95
0.07
0.38
348
33.5
66.4
7.88
31.7
6.75
2.18
6.80
1.06
6.19
1.22
3.32
0.47
3.12
0.46
6.02
3.04
0.73
0.06
2.44
4.28
1.28
0.702670±6
0.513075±18
8.6
18.5512±23
15.5314±28
38.1200±93
0.702883±7
0.512981±6
6.7
18.8472±38
15.5822±47
38.4282±154
0.702890±6
0.512980±5
6.6
18.8472±38
15.5838±29
38.4275±91
0.702880±4
0.512979±5
6.6
18.8498±23
15.5839±24
38.4319±75
0.702953±5
0.512960±5
6.2
18.9222±120
15.6020±103
38.5493±269
rock type
Lat. (°N)
Long. (°W)
WD (m)
87
Sr/86Sr
Nd/144Nd
εNd
206
Pb/204Pb
207
Pb/204Pb
208
Pb/204Pb
143
0.08
0.44
417
38.6
76.8
8.99
35.3
7.34
2.30
6.78
1.06
6.17
1.21
3.25
0.46
3.04
0.44
6.81
3.53
2.66
4.97
1.53
0.702900±4 0.703001±5
0.512993±6 0.512963±12
6.9
6.3
18.8392±17 18.9497±13
15.5824±21 15.6021±15
38.4407±69 38.5736±48
Errors for isotope data are 2SD and refer to the last digit.
0.04
0.19
207
19.6
41.5
5.15
21.8
5.13
1.69
5.14
0.83
5.03
1.00
2.72
0.39
2.51
0.37
4.36
1.78
50.81
4.02
12.50
13.33
0.20
4.88
9.26
2.76
0.85
0.46
0.03
0.03
0.05
99.15
2σ
BHVO2
n=5
0.64
0.15
0.53
0.34
0.06
0.25
0.16
0.16
0.05
0.09
0.03
0.11
0.11
0.91
4.64
31.7
317
284
45.1
121
129
106
21.5
9.31
392
26.2
175
18.0
4.06
1.70
0.13
0.10
130
15.3
37.7
5.34
24.7
6.08
2.06
6.14
0.95
5.30
0.97
2.44
0.32
2.00
0.28
4.33
1.12
0.26
0.02
1.67
1.19
0.41
Table A8: Geochemical and isotopic composition of the Galapagos Rise lavas.
sample
location
SO1603DS-1
SO160 3DS-2
SO160 6DS-1
SO160 3DS-3
SO160 8DS-3
SO160 8DS-5
SO160 7DS-1
SO160 8DS-1
SO160 10DS-1
SO160 10DS-5
Rift of Galapagos Rift of Galapagos Rift of Galapagos Rift of Galapagos Rift of Galapagos Rift of Galapagos Rift of Galapagos Rift of Galapagos Volcanic ridge, N Volcanic ridge, N
rift zone
Rise
Rise
Rise
Rise
Rise
Rise
Rise
Rise
rift zone
Lat. (°N)
-9.387
-9.387
-9.680
-9.387
-9.817
-9.817
-9.763
-9.817
-10.206
-10.206
Long. (°W)
WD (m)
94.406
3896
94.406
3896
94.427
3814
94.406
3896
94.424
2973
94.424
2973
94.422
3432
94.424
2973
94.502
2554
94.502
2554
Age (Ma)
Material
9
Glass
Glass
Glass
Whole rock
Whole rock
Whole rock
Whole rock
Whole rock
Glass
Glass
SiO2
TiO2
51.17
2.02
50.97
2.03
51.22
1.17
51.42
1.87
50.28
1.22
51.33
1.21
49.54
1.99
50.05
1.01
49.13
2.06
49.07
2.07
Al2O3
14.39
10.97
14.42
10.89
15.33
8.51
15.57
8.17
15.61
8.45
15.11
7.95
14.74
10.33
15.59
8.40
17.77
8.14
17.74
8.23
0.17
6.22
11.97
2.76
0.38
0.18
0.15
8.00
12.17
2.47
0.13
0.13
0.16
8.12
12.67
2.50
0.14
0.09
0.17
7.24
11.07
2.65
0.28
0.23
0.15
8.61
12.60
2.32
0.10
0.09
0.16
6.58
10.01
3.39
1.20
0.49
0.12
0.06
0.14
6.51
9.97
3.40
1.21
0.49
0.11
0.06
0.85
100.40
0.04
100.21
0.33
99.72
0.15
100.01
99.09
99.00
67.91
50.91
2.43
7.58
59.23
48.82
2.43
8.24
67.99
49.65
2.26
8.04
62.63
48.09
3.26
6.88
62.12
48.06
3.26
6.85
5.87
45.6
442
55.9
41.4
124
57.7
91.9
6.98
37.8
264
386
41.0
104.0
80.3
64.3
6.12
27.3
304
161
34.2
122
32.4
65.4
5.32
25.9
287
158
33.2
123
30.7
64.2
3.78
121
45.0
142.0
3.18
0.32
1.51
0.06
1.32
103
23.4
56
0.6
0.11
0.73
0.04
28.5
290
29.5
167
35.4
1.72
1.33
0.10
28.1
288
29.0
163
34.3
1.63
1.32
0.05
(wt. %)
FeOT
MnO
MgO
CaO
Na2O
K2 O
P2 O 5
SO2
Cl
LOI
Total
0.19
6.82
11.18
2.93
0.12
0.29
0.16
0.00
0.18
6.84
11.13
3.00
0.12
0.29
0.16
0.00
0.17
8.34
12.92
1.99
0.29
0.18
0.27
0.01
100.24
100.02
100.39
0.47
100.09
Mg#
Si72
Na72
Fe72
56.32
50.79
2.59
8.02
56.57
50.57
2.66
8.01
68.84
51.02
1.93
8.25
61.21
50.47
2.60
6.59
Li
Sc
V
Cr
Co
Ni
Cu
Zn
Ga
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
7.22
37.4
352
209
38.8
74.5
58.5
87.0
7.53
37.6
353
215
39.1
75.1
56.7
86.2
1.54
111
41.6
131
2.44
0.29
1.30
0.01
2.55
30.8
237
222
39.6
72.7
49.3
80.6
16.5
4.70
148
41.0
120
2.51
0.32
1.33
0.03
5.22
37.1
277
286
39.8
106
69.8
65.3
1.72
110
41.8
132
2.48
0.35
1.29
0.03
4.68
43.1
279
358
43.1
123
85.2
64.0
15.6
4.34
125.4
23.9
72.4
6.99
0.66
0.62
0.03
1.56
129
26.7
75.1
1.26
0.30
0.92
0.01
2.11
31.6
194
293
36.6
70.9
60.0
61.0
14.0
2.40
108
28.8
63.0
0.8
0.10
0.79
0.02
Cs
Ba
La
0.03
15.7
4.14
0.02
14.6
4.10
0.04
54.5
5.52
0.14
40.1
4.36
0.04
6.73
2.38
0.11
14.1
2.00
0.12
24.4
4.84
0.06
7.32
1.63
0.28
294
21.1
0.28
288
20.6
(ppm)
Ce
13.4
13.3
13.2
14.0
7.99
7.06
15.3
5.73
42.7
41.9
Pr
Nd
Sm
2.42
13.4
4.77
2.40
13.3
4.73
1.93
8.62
2.61
2.52
13.7
4.92
1.41
7.80
2.87
1.37
7.95
3.07
2.63
14.30
4.96
1.07
6.08
2.32
5.39
22.6
5.41
5.25
22.0
5.24
Eu
1.61
1.61
0.95
1.65
1.06
1.17
1.62
0.90
1.74
1.69
Gd
Tb
Dy
6.13
1.16
7.84
6.17
1.16
7.81
3.36
0.61
3.94
6.43
1.22
8.19
3.81
0.72
4.78
4.36
0.84
5.75
6.44
1.19
7.96
3.24
0.62
4.18
5.60
0.94
5.82
5.45
0.91
5.60
Ho
Er
1.67
4.78
1.66
4.73
0.86
2.45
1.75
4.99
1.02
2.89
1.23
3.51
1.69
4.81
0.91
2.57
1.18
3.29
1.14
3.15
Tm
Yb
0.69
4.59
0.70
4.62
0.36
2.42
0.74
4.86
0.43
2.83
0.52
3.45
0.71
4.68
0.38
2.51
0.47
3.09
0.45
3.01
Lu
Hf
0.67
3.47
0.68
3.46
0.36
1.85
0.73
3.90
0.41
2.17
0.52
2.25
0.69
4.01
0.37
1.70
0.46
3.86
0.44
3.71
Ta
Pb
0.16
0.61
0.16
0.54
0.39
0.58
0.21
1.03
0.08
0.71
0.07
0.56
0.22
0.75
0.06
0.44
1.85
1.44
1.76
1.42
Th
U
0.16
0.06
0.16
0.06
0.46
0.15
0.20
0.10
0.10
0.04
0.05
0.02
0.25
0.13
0.05
0.03
2.54
0.64
2.45
0.62
0.70258
0.70262
0.70252
0.70301
0.51316
0.51320
0.51300
18.089
15.476
37.680
17.968
15.437
37.420
18.788
15.541
38.552
Sr/86Sr
Nd/144Nd
0.70251
0.70293
0.70264
0.51321
0.51308
0.51320
Pb/ Pb
Pb/204Pb
208
Pb/204Pb
17.920
15.459
37.554
18.657
15.609
38.553
17.947
15.468
37.522
87
143
206
207
204
17.921
15.461
37.419
Table A8: Geochemical and isotopic composition of the Galapagos Rise lavas.
sample
location
SO160 11DS-1
SO160 11DS-2
SO160 12DS-1
SO160 14DS-1
SO160 14DS-2
SO160 14DS-4
Volcanic ridge, N Volcanic ridge, N Volcanic ridge, W Volcanic ridge, W Volcanic ridge, W Volcanic ridge, W
rift zone
rift zone
flank
flank
flank
flank
SO160 15DS-1
SO160 16DS-1
SO160 18DS-1
SO160 19DS-1
Volcanic ridge,
top
Volcanic ridge,
SW flank
Volcanic ridge,
SW flank
Volcanic ridge
Lat. (°N)
-10.262
-10.262
-10.255
-10.378
-10.378
-10.378
-10.423
-10.487
-10.6836
-10.9858
Long. (°W)
WD (m)
94.549
1922
94.549
1922
94.676
3520
94.648
2320
94.648
2320
94.648
2320
94.584
654
94.6872
2182
94.764
2588
94.8968
2526
Whole rock
Whole rock
Glass
Glass
6
Glass
Glass
Whole rock
Whole rock
Glass
7.5
Glass
SiO2
TiO2
47.90
2.14
47.71
2.06
48.91
1.95
50.62
1.65
50.35
1.81
50.17
1.71
44.31
1.67
52.94
1.74
50.61
2.28
48.91
1.69
Al2O3
18.38
9.36
18.13
9.26
17.48
7.99
16.50
9.27
15.75
9.36
16.86
8.88
18.77
9.71
18.58
6.51
18.34
6.84
17.04
8.31
0.14
6.57
10.03
3.44
1.25
0.47
0.12
0.06
0.17
6.87
8.63
3.56
1.59
0.53
0.04
0.07
0.18
5.79
10.68
3.35
1.09
0.40
0.07
0.06
0.17
5.57
9.61
3.65
1.49
0.53
0.04
0.06
0.15
6.41
9.08
3.06
1.00
1.74
0.15
3.43
5.50
5.41
2.68
0.80
0.12
4.73
7.93
4.49
2.28
0.75
0.10
0.12
0.13
7.51
11.10
2.97
0.82
0.40
0.12
0.05
98.89
98.73
2.63
99.61
1.25
99.71
98.59
99.05
58.88
49.93
4.25
4.66
65.20
47.97
2.89
7.54
Age (Ma)
Material
(wt. %)
FeOT
MnO
MgO
CaO
Na2O
K2 O
P2 O 5
SO2
Cl
LOI
Total
0.23
5.22
9.89
3.64
1.32
0.56
0.17
5.95
9.95
3.42
1.23
0.58
1.82
100.46
1.87
100.33
98.41
Mg#
Si72
Na72
Fe72
53.62
47.65
3.18
5.48
57.12
47.25
3.11
6.56
63.03
47.86
3.32
6.81
60.56
49.74
3.38
7.53
56.21
49.98
3.00
6.37
56.54
49.77
3.32
5.99
Li
Sc
V
Cr
Co
Ni
Cu
Zn
Ga
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
10.9
27.5
306
173
52.4
124
37.3
99.9
10.2
26.7
286
178
37.0
127
34.7
81.2
5.11
26.2
285
197
31.9
119
34.3
63.3
7.40
22.3
259
211
32.3
134
36.9
72.2
5.37
32.3
302
163
31.4
62
41.8
64.5
6.34
23.2
259
269
34.5
153
39.1
69.7
12.60
29.3
316
108
35.1
98.5
25.0
78.3
9.25
11.5
143
34.4
20.3
44.5
16.6
70.4
5.97
20.5
233
68.7
21.6
42.9
23.6
54.2
5.38
30.3
300
268
37.7
144
49.1
66.9
18.2
411
34.7
185
37.8
2.22
1.67
1.15
14.9
410
29.8
180
36.7
1.53
1.57
0.51
29.7
287
28.5
164
35.9
1.78
1.31
0.04
32.1
335
30.0
175
36.3
1.97
1.18
0.09
19.8
287
19.8
117
21.8
1.11
1.00
0.05
28.7
320
28.1
155
32.4
1.77
1.11
0.08
11.1
31.4
163
25.0
1.04
0.89
1.59
55.7
594
31.7
347
77.9
2.71
2.32
0.22
59.1
355
33.5
272
69.6
3.5
1.96
0.09
17.2
238
26.7
132
23.8
1.20
1.07
0.04
Cs
Ba
La
0.19
318
25.3
0.19
306
23.8
0.29
302
21.0
0.32
421
28.8
0.19
234
14.5
0.29
379
25.4
0.03
388
27.7
0.63
665
53.8
0.59
510
39.3
0.17
170
14.6
(ppm)
Ce
49.4
47.7
41.9
54.7
29.5
48.5
52.9
96
73.4
30.9
Pr
Nd
Sm
6.02
24.8
5.66
5.76
23.6
5.43
5.17
21.4
5.07
6.44
25.4
5.39
3.73
15.9
3.92
5.69
22.7
4.96
6.17
24.3
5.28
10.4
38.1
7.17
8.40
32.4
6.76
4.02
17.3
4.41
Eu
1.83
1.74
1.62
1.70
1.34
1.55
1.68
2.19
2.04
1.47
Gd
Tb
Dy
5.79
0.94
5.79
5.51
0.89
5.49
5.19
0.88
5.42
5.44
0.91
5.69
4.28
0.74
4.75
4.93
0.84
5.23
5.41
0.89
5.58
6.35
0.97
5.62
6.53
1.05
6.35
4.77
0.82
5.24
Ho
Er
1.19
3.32
1.10
3.03
1.11
3.09
1.17
3.32
0.99
2.76
1.08
3.06
1.16
3.25
1.11
3.06
1.27
3.58
1.07
2.99
Tm
Yb
0.48
3.23
0.44
2.92
0.44
2.96
0.49
3.35
0.40
2.69
0.45
3.04
0.48
3.23
0.45
3.04
0.52
3.49
0.43
2.88
Lu
Hf
0.49
4.50
0.43
4.37
0.44
3.61
0.51
3.94
0.40
2.70
0.46
3.43
0.48
3.97
0.47
8.16
0.52
5.63
0.42
3.14
Ta
Pb
2.26
1.94
2.21
1.83
1.81
1.50
1.87
2.17
1.12
1.31
1.60
1.96
1.36
2.06
4.92
5.1
3.48
2.7
1.28
1.04
Th
U
3.25
1.07
3.14
0.94
2.60
0.67
3.42
0.85
1.67
0.43
3.09
0.77
3.74
1.73
8.88
1.99
5.73
1.43
1.58
0.43
87
Sr/86Sr
Nd/144Nd
143
Pb/ Pb
Pb/204Pb
208
Pb/204Pb
206
207
204
0.70311
0.70303
0.70304
0.70296
0.51297
18.500
0.51302
18.917
0.51299
18.812
0.51301
18.981
15.532
38.322
15.581
38.677
15.557
38.587
15.540
38.574
Table A8: Geochemical and isotopic composition of the Galapagos Rise lavas.
sample
SO160 19DS-4
SO160 20DS-1
SO160 21DS-1
location
Volcanic ridge
Volcanic ridge
Volcanic ridge
Lat. (°N)
-10.9858
-11.146
-11.1533
Long. (°W)
WD (m)
94.8968
2526
94.9624
2507
94.9508
2011
Glass
Glass
Glass
SiO2
TiO2
48.86
1.68
48.54
2.00
49.78
2.02
Al2O3
17.07
8.31
17.58
7.99
16.69
8.81
0.16
7.48
11.06
2.93
0.81
0.39
0.11
0.05
0.14
7.12
10.46
3.37
1.06
0.46
0.10
0.05
0.17
5.98
9.63
3.36
1.16
0.44
0.07
0.06
98.89
98.87
98.19
65.10
47.92
2.84
7.52
64.88
47.57
3.29
7.16
58.48
49.16
3.11
6.51
Li
Sc
V
Cr
Co
Ni
Cu
Zn
Ga
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
5.18
29.6
293
269
38.5
160
48.7
65.4
5.10
26.8
279
172
33.9
116
36.9
62.8
6.99
29.7
358
112
33.4
121
29
73.8
16.7
234
26.0
129
23.3
1.17
1.14
0.04
21.1
293
28.5
169
31.2
1.59
1.34
0.05
25.1
281
30.6
161
29.4
1.48
1.31
0.06
Cs
Ba
La
0.17
166
14.2
0.21
208
18.7
0.25
293
19.3
Age (Ma)
Material
(wt. %)
FeOT
MnO
MgO
CaO
Na2O
K2 O
P2 O 5
SO2
Cl
LOI
Total
Mg#
Si72
Na72
Fe72
(ppm)
Ce
30.2
39.1
40.2
Pr
Nd
Sm
3.90
16.85
4.28
5.02
21.5
5.20
5.07
21.6
5.23
Eu
1.42
1.69
1.71
Gd
Tb
Dy
4.65
0.80
5.07
5.42
0.91
5.58
5.50
0.94
5.92
Ho
Er
1.03
2.92
1.13
3.13
1.22
3.41
Tm
Yb
0.42
2.78
0.45
2.95
0.50
3.29
Lu
Hf
0.41
3.03
0.43
3.87
0.49
3.69
Ta
Pb
1.25
1.11
1.66
1.24
1.52
1.56
Th
U
1.55
0.43
2.1
0.57
2.18
0.56
0.70291
0.70307
0.51303
18.918
0.51300
18.616
15.530
38.530
15.521
38.394
87
Sr/86Sr
Nd/144Nd
143
Pb/ Pb
Pb/204Pb
208
Pb/204Pb
206
207
204
Table A9: Major and trace element and radiogenic isotope data of lavas from the 8°48'S volcanic field.
sample
location
Lat. (°S)
M41/2 156 DS-1 M41/2 156 DS-3 M41/2 156 DS-4
Old lavas
Old lavas
Old lavas
M64/1 159ROV6
M64/1 159ROV5
M64/1 159ROV2 M64/1 155ROV-1 avge M64/1 148VSR
Old pillow mound Old pillow mound
Old lavas
Old lavas
Old lavas
8.749
8.749
8.749
8.797
8.799
8.803
8.816
8.817
13.503
13.503
13.503
13.503
13.503
13.502
13.508
13.497
2257
2257
2257
2151
2186
2201
2161
2230
SiO2
50.37
50.17
50.38
51.29
49.70
51.17
50.65
50.55
TiO2
1.24
1.16
1.24
1.46
1.24
1.17
1.31
1.53
Al2O3
16.05
15.65
16.08
15.33
15.42
14.86
15.74
15.15
FeO
MnO
10.61
10.51
10.36
9.62
10.15
9.49
10.04
10.10
0.18
0.18
0.18
0.15
0.18
0.17
0.15
0.17
MgO
7.93
7.86
7.84
7.66
7.93
8.19
7.93
7.73
CaO
10.73
11.92
10.36
10.48
11.68
12.39
10.65
11.13
Na2O
2.61
2.49
2.67
2.53
2.39
2.08
2.56
2.49
K2 O
0.09
0.11
0.13
0.10
0.14
0.11
0.06
0.13
P2O5
0.14
0.18
0.23
0.20
0.21
0.20
0.19
0.22
S (ppm)
910
1091
1080
850
1055
1187
1563
1420
Long. (°W)
WD (m)
(wt. %)
T
Cl (ppm)
77
129
100
34
101
307
444
381
Total
100.41
100.40
99.55
98.92
99.16
99.93
99.37
99.34
Mg#
62.3
61.8
64.2
62.1
61.3
59.4
48.7
52.0
Li
2.46
3.70
3.77
4.28
Sc
28.8
30.3
31.0
30.8
V
168
189
198
199
Cr
284
311
323
343
Co
43.8
45.6
48.8
48.8
177
(ppm)
Ni
158
154
170
Cu
97.9
95.4
96.6
112
Zn
73.9
85.3
82.9
89.3
Ga
14.8
16.1
16.0
16.3
Rb
1.40
1.07
1.12
0.82
Sr
138
158
152
142
Y
25.0
17.6
18.1
19.6
Zr
65.7
47.5
49.1
61.2
Nb
3.67
2.70
2.82
2.94
Mo
0.28
Sn
0.82
Sb
0.01
Cs
0.02
0.02
0.01
0.03
0.01
Ba
18.7
14.1
14.2
10.6
La
2.89
2.11
2.17
1.82
Ce
7.84
5.83
5.96
5.27
Pr
1.28
0.99
1.01
0.96
Nd
6.75
5.43
5.57
5.58
Sm
2.47
2.04
2.05
2.28
Eu
1.05
0.87
0.88
0.97
Gd
3.39
2.80
2.83
3.20
Tb
0.61
0.51
0.52
0.58
Dy
4.03
3.31
3.39
3.67
Ho
0.83
0.69
0.70
0.75
Er
2.30
1.85
1.91
1.96
Tm
0.34
0.28
0.28
0.28
Yb
2.21
1.76
1.81
1.82
Lu
0.33
0.26
0.27
0.27
Hf
1.76
1.41
1.44
1.79
Ta
0.24
0.16
0.16
0.20
Pb
0.22
0.20
0.23
0.17
Th
0.20
0.14
0.13
0.10
U
0.06
0.06
0.05
0.04
W
87
0.06
Sr/86Sr
0.70228
0.70239
143
Nd/144Nd
0.51319
0.51322
206
Pb/204Pb
18.736
18.900
207
Pb/204Pb
Pb/204Pb
15.543
15.552
208
234
(
38.202
38.257
U/238U
230
Th/
232
1.006
Th)
(238U/232Th)
(
230
(
226
Th/
238
U)
230
Ra/
Th)
1.078
1.140
1.131
1.141
1.028
1.079
0.940
1.051
1.049
1.057
1.203
1.086
1.036
0.924
1.415
0.978
Trace element and isotope data have been previously published by Hoernle et al. (2011).
Table A9: Major and trace element and radiogenic isotope data of lavas from the 8°48'S volcanic field.
sample
location
M64/1 159ROV10 M64/1 159ROV11 M64/1 159ROV9 M64/1 157VSR M64/1 159ROV7 M64/1 159ROV8 M64/1 159ROV4 M64/1 152VSR avge
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Lat. (°S)
8.791
8.791
8.792
8.795
8.796
8.796
8.800
8.800
Long. (°W)
13.503
13.503
13.504
13.509
13.504
13.504
13.502
13.497
2219
2219
2215
2190
2201
2202
2201
2223
SiO2
49.59
50.13
50.37
51.72
50.88
50.91
50.83
50.59
TiO2
1.44
2.36
2.12
1.52
2.00
1.92
1.37
1.53
Al2O3
14.50
13.66
13.98
14.38
13.67
13.85
15.16
15.20
FeO
MnO
10.38
12.17
11.25
10.16
11.38
10.89
8.78
9.06
0.17
0.22
0.21
0.21
0.20
0.19
0.16
0.15
MgO
7.33
5.57
5.89
7.29
6.11
6.59
8.14
7.91
CaO
12.36
10.10
10.61
11.59
10.57
11.18
12.42
11.96
Na2O
2.41
3.14
2.85
2.29
2.63
2.54
2.33
2.36
K2 O
0.19
0.41
0.41
0.23
0.27
0.27
0.21
0.22
P2O5
0.24
0.38
0.36
0.25
0.31
0.29
0.26
0.25
S (ppm)
1197
1437
1382
997
991
980
1116
1129
WD (m)
(wt. %)
T
Cl (ppm)
142
177
184
149
114
157
136
114
Total
98.76
98.34
98.23
99.75
98.19
98.79
99.76
99.35
Mg#
59.8
52.7
55.6
65.8
64.4
65.5
62.9
62.3
Li
5.87
4.31
6.36
5.06
4.29
4.68
Sc
41.4
36.7
39.5
42.9
36.8
35.9
V
284
232
337
303
235
243
Cr
217
365
163
476
361
334
Co
42.2
42.7
49.2
58.3
42.3
42.8
Ni
77.7
136
75.4
182
128
133
Cu
88.9
88.5
89.6
122
88.5
91.6
(ppm)
Zn
90.1
71.7
107
94.0
68.9
74.9
Ga
17.0
14.6
20.1
19.8
14.7
15.9
Rb
3.35
3.42
4.42
4.29
3.41
3.46
Sr
164
168
169
191
167
186
Y
27.5
19.5
32.9
23.7
19.7
22.5
Zr
91.3
72.9
118
82.7
73.4
89.1
Nb
8.56
7.52
11.1
8.53
7.54
8.49
Mo
0.68
0.55
Sn
1.04
0.40
Sb
0.02
0.01
Cs
0.04
0.04
0.05
0.05
0.04
0.04
Ba
42.2
44.4
50.7
43.1
44.4
41.4
La
5.72
4.75
6.82
4.75
4.73
5.16
Ce
14.5
11.5
16.9
11.4
11.6
12.9
Pr
2.21
1.77
2.61
1.73
1.76
1.97
Nd
114
8.74
13.0
8.47
8.90
10.0
Sm
3.57
2.73
3.94
2.60
2.73
3.13
Eu
1.31
1.03
1.38
0.97
1.02
1.17
Gd
4.42
3.34
4.72
3.15
3.35
3.80
Tb
0.76
0.58
0.82
0.54
0.58
0.65
Dy
4.94
3.70
5.27
3.42
3.69
4.11
Ho
1.00
0.75
1.08
0.69
0.75
0.83
Er
2.75
2.04
2.97
1.88
2.04
2.20
Tm
0.40
0.30
0.44
0.28
0.30
0.32
Yb
2.58
1.94
2.90
1.79
1.93
2.09
Lu
0.38
0.29
0.43
0.26
0.28
0.30
Hf
2.60
1.93
2.80
1.73
1.92
2.27
Ta
0.49
0.43
0.67
0.44
0.48
0.49
0.15
0.10
W
0.11
Pb
0.44
0.40
0.68
0.35
0.37
0.46
Th
0.44
0.39
0.52
0.35
0.38
0.39
U
0.13
0.13
0.17
0.12
0.12
0.13
87
Sr/86Sr
0.70251
0.70247
143
Nd/144Nd
0.51316
0.51320
206
Pb/204Pb
19.005
18.903
207
Pb/204Pb
Pb/204Pb
15.574
15.565
38.579
38.422
208
234
U/238U
(
230
(
238
(
230
(
226
Th/
232
232
U/
Th/
0.993
Th)
Th)
238
U)
230
Ra/
Th)
1.117
1.146
1.130
1.152
1.127
1.093
0.913
0.938
0.934
0.949
0.931
0.925
1.117
1.222
1.210
1.215
1.210
1.181
1.466
1.400
0.972
Trace element and isotope data have been previously published by Hoernle et al. (2011).
1.493
Table A9: Major and trace element and radiogenic isotope data of lavas from the 8°48'S volcanic field.
sample
location
M64/1 151VSR avge M64/1 159ROV3 M64/1 159ROV1
Southern lavas
Southern lavas
Southern lavas
M64/1 156VSR
M41/2 157 DS-1
Southern lavas
Southern lavas
M41/2 157 DS-2 M41/2 157 DS-3 M64/1 155ROV-4
Southern lavas
Southern lavas
Southern lavas
Lat. (°S)
8.800
8.801
8.803
8.807
8.808
8.808
8.808
8.816
Long. (°W)
13.502
13.502
13.502
13.507
13.496
13.496
13.496
13.503
2219
2198
2204
2208
2212
2212
2212
2195
SiO2
50.52
51.15
51.47
51.37
50.82
50.79
50.97
50.97
TiO2
1.37
1.46
1.24
1.66
1.50
1.50
1.49
1.34
Al2O3
15.09
14.69
14.46
14.33
15.52
15.77
15.82
15.40
FeO
MnO
8.75
9.38
9.60
10.25
9.04
8.97
8.98
8.61
0.16
0.19
0.17
0.18
0.14
0.15
0.15
0.16
MgO
8.03
7.66
7.66
7.20
8.10
8.09
8.15
8.19
CaO
12.38
12.21
12.18
11.67
12.00
11.87
11.81
12.44
Na2O
2.30
2.34
2.09
2.50
2.37
2.44
2.42
2.31
K2 O
0.23
0.21
0.19
0.22
0.22
0.25
0.23
0.22
P2O5
0.24
0.25
0.23
0.26
0.16
0.31
0.30
0.25
S (ppm)
1225
992
979
1011
1005
1212
WD (m)
(wt. %)
T
Cl (ppm)
143
104
133
155
144
173
Total
99.18
99.67
99.42
99.75
100.12
100.22
100.37
100.01
Mg#
59.3
65.0
65.2
65.3
66.3
65.7
56.9
65.5
(ppm)
Li
4.39
2.36
4.23
Sc
36.2
31.6
36.4
V
244
198
235
Cr
334
335
361
Co
44.5
38.1
42.3
Ni
132
129
132
Cu
90.0
85.8
87.6
Zn
79.8
62.8
78.9
Ga
15.6
13.9
14.6
Rb
3.04
3.12
3.27
Sr
169
180
166
Y
20.7
24.6
19.6
Zr
77.4
90.1
72.3
Nb
7.14
7.86
7.34
Mo
0.51
Sn
0.70
Sb
0.01
0.02
Cs
0.04
0.04
0.04
Ba
39.7
51.2
42.5
La
4.55
5.29
4.63
Ce
11.4
13.6
11.4
Pr
1.77
2.09
1.73
Nd
8.92
10.3
8.71
Sm
2.84
3.25
2.69
Eu
1.07
1.25
1.02
Gd
3.53
3.96
3.31
Tb
0.61
0.67
0.58
Dy
3.88
4.21
3.67
Ho
0.79
0.83
0.75
Er
2.13
2.26
2.02
Tm
0.31
0.32
0.30
Yb
2.02
2.10
1.9
Lu
0.30
0.31
0.28
Hf
2.02
2.30
1.93
Ta
0.40
0.49
0.41
Pb
0.35
0.41
0.36
Th
0.34
0.39
0.37
U
0.11
0.12
0.12
W
87
0.14
Sr/86Sr
0.702463
0.70248
0.70247
0.70247
143
Nd/144Nd
0.513199
0.51318
0.51320
0.51319
206
Pb/204Pb
18.9082
18.959
18.932
18.900
Pb/204Pb
208
Pb/204Pb
15.5687
15.571
15.563
15.561
38.4355
38.499
38.413
38.413
1.0071
0.999
1.001
1.009
0.998
(230Th/232Th)
1.621
1.157
1.148
1.190
1.180
(238U/232Th)
0.9311
0.940
0.941
0.960
0.941
(230Th/238U)
1.2481
1.231
1.220
1.240
1.180
0.997
1.310
207
234
U/238U
(226Ra/230Th)
Trace element and isotope data have been previously published by Hoernle et al. (2011).
Table A9: Major and trace element and radiogenic isotope data of lavas from the 8°48'S volcanic field.
sample
M64/1 155ROV-7
M64/1 155ROV-5
M64/1 155ROV-6
M64/1 155ROV-3
M64/1 155ROV-8
M64/1 165VSR
location
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Southern lavas
Lat. (°S)
M41/2 158 DS-1 M41/2 158 DS-2
Southern lavas
Southern lavas
M64/1 166VSR
Southern lavas
8.817
8.817
8.817
8.817
8.817
8.833
8.838
8.838
8.842
13.500
13.501
13.501
13.505
13.498
13.495
13.495
13.495
13.491
2221
2199
2190
2149
2218
2225
2139
2139
2188
SiO2
50.83
50.87
51.24
50.73
50.72
49.66
51.24
50.83
50.37
TiO2
1.49
1.37
1.72
1.72
1.44
2.47
1.69
1.69
3.15
Al2O3
15.49
15.25
14.25
14.45
15.65
13.95
14.36
14.34
13.34
FeO
MnO
8.90
8.80
10.60
10.20
8.74
11.59
10.14
10.09
13.22
0.16
0.15
0.18
0.19
0.13
0.19
0.19
0.21
0.21
MgO
8.15
8.14
6.75
7.24
8.30
5.68
7.16
7.00
4.40
CaO
12.07
12.48
11.11
11.53
12.17
10.30
11.45
11.34
8.93
Na2O
2.37
2.31
2.68
2.51
2.31
3.17
2.55
2.74
3.55
K2 O
0.22
0.23
0.27
0.26
0.22
0.54
0.29
0.30
0.72
P2O5
0.26
0.25
0.29
0.28
0.25
0.43
0.28
0.32
0.54
S (ppm)
1185
961
1346
1190
1656
1203
1164
1214
Long. (°W)
WD (m)
(wt. %)
T
Cl (ppm)
147
119
942
123
773
353
403
399
Total
100.06
99.95
99.22
99.24
100.04
98.21
99.42
99.24
98.68
Mg#
59.6
66.3
50.4
59.4
59.0
40.8
59.0
58.4
58.2
(ppm)
Li
4.3
6.18
5.18
4.77
4.95
3.29
Sc
35.5
43.0
40.6
34.8
34.7
36.0
V
238
299
303
252
252
240
Cr
338
107
142
402
365
158
Co
43.3
42.0
43.9
51.8
47.3
38.6
Ni
140
57.6
67.9
188
149
58.2
Cu
88.4
83.5
79.8
107
99.3
79.4
Zn
77.7
91.0
85.6
86.2
81.6
74.2
Ga
15.2
17.1
16.5
18.6
17.24
14.9
Rb
3.21
3.94
4.00
3.41
3.51
3.72
Sr
179
169
166
192
174
169
Y
20.6
29
26.5
22.8
22.3
29.1
Zr
83.8
99.2
105
86.2
76.2
105
Nb
7.89
6.69
10.0
8.13
7.79
9.09
Mo
0.52
0.51
Sn
0.78
0.69
Sb
0.01
0.01
0.02
Cs
0.04
0.05
0.05
0.04
0.04
0.04
Ba
41.7
49.8
49.2
39.7
41.6
57.1
La
4.85
6.8
6.39
4.67
4.76
6.25
Ce
12.2
16.8
16.0
11.7
11.4
16.0
Pr
1.9
2.54
2.44
1.80
1.76
2.43
Nd
9.65
12.8
12.3
9.11
8.65
11.8
Sm
3.02
3.83
3.71
2.82
2.68
3.67
Eu
1.12
1.41
1.33
1.05
1.01
1.37
Gd
3.64
4.70
4.49
3.38
3.26
4.48
Tb
0.63
0.81
0.78
0.58
0.57
0.76
Dy
3.93
5.24
4.94
3.62
3.61
4.88
Ho
0.79
1.06
1.01
0.72
0.73
0.98
Er
2.12
2.94
2.76
1.94
1.98
2.70
Tm
0.31
0.43
0.41
0.28
0.29
0.39
Yb
1.98
2.81
2.66
1.82
1.89
2.56
Lu
0.29
0.41
0.40
0.27
0.28
0.38
Hf
2.21
2.76
2.72
1.95
1.83
2.59
Ta
0.48
0.56
0.58
0.47
0.46
0.56
0.10
0.10
W
Pb
0.36
0.53
0.60
0.36
0.40
0.45
Th
0.37
0.54
0.49
0.34
0.37
0.49
U
0.12
0.165
0.16
0.11
0.13
0.15
87
Sr/86Sr
0.70248
0.70247
143
Nd/144Nd
0.51321
0.51316
206
Pb/204Pb
18.934
18.983
Pb/204Pb
208
Pb/204Pb
15.562
15.570
207
234
U/238U
38.409
38.524
1.004
1.007
1.004
1.000
1.000
1.000
(230Th/232Th)
1.171
1.141
1.095
1.169
1.180
1.149
(238U/232Th)
0.962
0.954
0.930
0.944
0.965
0.948
(230Th/238U)
1.217
1.196
1.177
1.238
1.223
1.212
(226Ra/230Th)
1.401
1.456
0.982
1.459
1.479
1.400
Trace element and isotope data have been previously published by Hoernle et al. (2011).
Table A9: Major and trace element and radiogenic isotope data of lavas from the 8°48'S volcanic field.
sample
M64/1 160VSR
M64/1 161VSR
M64/1 162VSR
M64/1 163VSR
location
Northern lavas
Northern lavas
Northern lavas
Northern lavas
Lat. (°S)
8.782
8.778
8.770
8.757
Long. (°W)
13.507
13.510
13.511
13.512
2208
2266
2273
2287
SiO2
48.49
48.02
48.07
47.98
TiO2
2.30
2.44
2.48
2.52
Al2O3
15.54
15.60
15.58
15.59
FeO
MnO
10.26
10.46
10.52
10.52
0.17
0.17
0.15
0.15
MgO
7.13
7.07
7.08
6.92
CaO
9.73
9.49
9.62
9.64
Na2O
3.19
3.27
3.23
3.25
K2 O
0.67
0.75
0.76
0.77
P2O5
0.46
0.49
0.49
0.49
S (ppm)
1233
0.12
0.12
0.12
Cl (ppm)
392
403
399
392
Total
98.09
97.92
98.14
97.99
Mg#
57.7
WD (m)
(wt. %)
T
(ppm)
Li
6.12
5.57
5.12
Sc
38.8
42.4
39.6
V
311
280
283
Cr
164
235
282
Co
47.3
42.2
50.6
Ni
76.7
85.3
108
Cu
86.3
88.5
117
Zn
97.1
81.9
91.5
Ga
18.5
15.5
19.2
Rb
3.56
2.91
3.26
Sr
140
117
172
Y
27.2
23.7
24.6
Zr
85.5
64.9
84.4
Nb
8.16
6.73
7.76
Mo
0.53
0.51
Sn
0.76
0.74
Sb
0.02
Cs
0.04
0.04
0.04
Ba
40.8
35.7
37.5
La
5.17
4.46
4.75
Ce
12.8
11.1
11.7
Pr
1.96
1.70
1.83
Nd
9.81
8.72
9.11
Sm
3.06
2.74
2.82
Eu
1.12
1.02
1.05
Gd
3.78
3.48
3.44
Tb
0.67
0.63
0.60
Dy
4.34
4.16
3.83
Ho
0.90
0.86
0.78
Er
2.49
2.41
2.14
Tm
0.37
0.36
0.31
Yb
2.43
2.36
2.08
Lu
0.36
0.35
0.31
Hf
2.08
1.92
1.91
Ta
0.49
0.39
0.45
W
0.12
Pb
0.52
0.42
0.39
Th
0.38
0.35
0.35
U
0.13
0.11
0.13
87
Sr/86Sr
0.02
0.10
0.70249
0.70238
0.70251
143
Nd/144Nd
0.51318
0.51325
0.51317
206
Pb/204Pb
18.988
18.565
18.994
Pb/204Pb
208
Pb/204Pb
15.572
15.528
15.573
38.515
38.109
38.544
207
234
U/238U
1.015
1.001
1.000
0.991
(230Th/232Th)
1.151
1.142
1.133
1.142
(238U/232Th)
0.959
0.979
0.956
0.963
(230Th/238U)
1.201
1.166
1.186
1.186
(226Ra/230Th)
1.650
1.769
Trace element and isotope data have been previously published by Hoernle et al. (2011).

Download Report

Dissertation Philipp A. Brandl

Paperzz.com

Your Paperzz