Continental transpressional and transtensional tectonics

Continental Transpressional and TranstensionaI Tectonics

Geological Society Special Publications Series Editors A. J. FLEET A. C. MORTON A. M. ROBERTS

It is recommended that reference to all or part of this book should be made in one of the following ways.

HOLDSWORTH,R. E., STRACHAN,R. A. & DEWEY,J. E (eds) 1998. Continental Transpressional and Transtensional Tectonics. Geological Society, London, Special Publications, 135.

LIN, S., JIANG,D. & WILLIAMS,P. E 1998. Transpression (-transtension) zones of triclinic symmetry: natural example and theoretical modelling. In: HOLDSWORTH,R. E., STRACHAN,R. A. & DEWEY,J. E (eds) Continental Transpressional and Transtensional Tectonics. Geological Society, London, Special Publications, 135, 41-58.

G E O L O G I C A L S O C I E T Y S P E C I A L P U B L I C A T I O N NO. 135

Continental Transpressional and Transtensional Tectonics EDITED

BY

R. E. H O L D S W O R T H University of Durham, UK R. A. S T R A C H A N Oxford Brookes University, UK AND J. F. D E W E Y University of Oxford, UK

1998 Published by The Geological Society London

THE GEOLOGICAL SOCIETY

The Society was founded in 1807 as The Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of around 8000. It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists, SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' relevant experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London W l V 0JU, UK. The Society is a Registered Charity, No. 210161. Published by The Geological Society from: The Geological Society Publishing House Unit 7, Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel. 01225 445046 Fax 01225 442836)

Distributors

USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders: Tel. (918) 584-2555 Fax (918) 560-2652)

First published 1998

Australia The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. 9 The Geological Society 1998. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/98/$10.00.

Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5065 Australia (Orders: Tel. (08) 379-0444 Fax (08) 379-4634)

India Affiliated East-West Press PVT Ltd G-1/16 Ansari Road New Delhi 110 002 India (Orders: Tel. (11) 327-9113 Fax (11) 326--0538)

Japan British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN1-86239-007-X Typeset by Type Study, Scarborough, UK Printed in Great Britain by The Alden Press, Osney Mead, Oxford, UK.

Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surugadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel. (03) 3255-3497 Fax (03) 3255-3495)

Contents

Preface Acknowledgements

DEWEY, J. E, HOLDSWORTH,R. E. & STRACHAN,R. A. Transpression and transtension

xii xiii 1

zones

Modelling Transpression and Transtension FOSSEN, H. & TIKOFF, B. Extended models of transpression and transtension, and application to tectonic settings

15

JONES, R. R. & HOLDSWORTH,R. E. Oblique simple shear in transpression zones

35

LIN, S., JIANG, D. & WILLIAMS,P. E Transpression (or transtension) zones of triclinic symmetry: natural example and theoretical modelling

41

SCHREURS, G. & COLLETTA,g. Analogue modelling of faulting in zones of continental transpression and transtension

59

Continental Transform Zones BUTLER, R. W. H., SPENCER,S. & GRIFFITHS,H. M. The structural response to evolving

81

plate kinematics during transpression: evolution of the Lebanese restraining bend of the Dead Sea Transform TAVARNELLI,E. Tectonic evolution of the Northern Salinian Block, California, USA: Paleogene to Recent shortening in a transform fault-bounded continental fragment

107

RUST, D. Contractional and extensional structures in the transpressive 'Big Bend' of the San Andreas fault, southern California

119

REIJS, J. & MCCLAY, K. Salar Grande pull-apart basin, Atacama Fault System, northern Chile

127

TEYSSIER, C. & TIKOFF,B. Strike-slip partitioned transpression of the San Andreas fault system: a lithospheric-scale approach

143

Oblique Divergence Zones KRABBENDAM,M. & DEWEY,J. F. Exhumation of U H P rocks by transtension in the Western Gneiss Region, Scandanavian Caledonides

159

DOKKA, R. K., ROSS, T. M. & Lu, G. The Trans Mojave-Sierran shear zone and its role in Early Miocene collapse of southwestern North America

183

WATKEYS,M. K. & SOKOUTIS,D. Transtension in southeastern Africa associated with

203

Gondwana break-up ALLEN, M. B., MACDONALD,D. I. M., ZHAO NUN.,VINCENT,S. J. & BROUET-MENZIES,C. Transtensional deformation in the evolution of the Bohai Basin, northern China

215

Oblique Convergence Zones EBERT, H. D. & HASUI,Y. Transpressional tectonics and strain partitioning during oblique collision between three plates in the Precambrian of southeast Brazil

231

GAYER,R., HATHAWAY,Z. & NEMCOK,M. Transpressionally driven rotation in the external orogenic zones of the Western Carpathians and the SW British Variscides

253

GLEIZES, G., LEBLANC,D. & BOUCHEZ,J. L. The main phase of the Hercynian Orogeny in the Pyrenees is a dextral transpression

267

TANNER, D. C., BEHRMANN, J. H., ONCKEN, O. & WEBER, K. Three-dimensional retro-modelling of transpression on a linked fault system: the Upper Cretaceous deformation on the western border of the Bohemian Massif, Germany

275

CURTIS,M. L. Development of kinematic partitioning within a pure-shear dominated

289

dextral transpression zone: the southern Ellsworth Mountains, Antarctica

SEARLE,M. R, WEINBERG,R. E & DUNLAP,W. J. Transpressional tectonics along the Karakoram Fault Zone, northern Ladakh: constraints on Tibetan extrusion

307

SAINT BLANQUAT,M., TIKOFF, B., TEYSSIER, C. • VIGNERESSE, J. L. Transpressional kinematics and magmatic arcs

327

SCHIATTARELLA,M. Quaternary tectonics of the Pollino Ridge, Calabria-Lucania boundary, southern Italy

341

Index

355

Preface

This volume contains a broad spectrum of papers that summarize recent advances in the understanding of continental transpressional and transtensional tectonics. The papers include theoretical and case studies from a global set of contributors. The volume contains 22 papers. The opening contribution by Dewey et al. is an overview of the basic features of transpressional and transtensional deformation zones aimed at setting the scene for the more detailed papers to follow. These are grouped into four sections. The first, Modelling Transpression and Transtension, includes a series of papers which discuss theoretical strain models in the context of field examples and analogue experiments (Fossen & Tikoff, Jones & Holdsworth, Lin et al. and Schreus & Colletta). The second section details the tectonic evolution of Continental Transform Zones and includes papers on the Dead Sea Transform (Buffer et al.), the San Francisco Bay area (Tavarnelli), the Transverse Ranges of southern California (Rust), the Atacama Fault System of Chile (Rjeiis & McClay) and a lithosphere-scale view of the San Andreas fault system (Teyssier & Tikoff). The third section is entitled Oblique Divergence Zones. The first two papers are concerned with transtensional structures developed during gravitational collapse in the Caledonides of western Norway (Krabbendam & Dewey) and in southwestern North America (Dokka et al.). The two following papers describe the transtensional structures developed in South Africa during break-up of Gondwana (Watkeys & Soukoutis) and the evolution of the combined pull-apart/transtensional Bonai Basin, northern China (AHen et al.). The fourth section is concerned with Oblique Convergence Zones and includes case studies from the Precambrian basement of Brazil (Ebert & Hasui), the European Variscides (Gayer et al., Gleizes el al. and Tanner et al.), the Permo-Triassic Gondwanian orogen of west Antarctica (Curtis), the Himalayas (Searle et al.) and the Sierra Nevada batholith (St Blanquat et al.), and concludes with a discussion of Quaternary tectonics in southern Italy (Shiatterella). The impetus for this volume was a conference held in March 1997 at Burlington House, London, under the auspices of the Tectonic Studies Group of the Geological Society of London. The editors would like to thank all the staff at Burlington House who helped ensure the smooth running of the conference, including projection facilities and refreshments. Bob Holdsworth, Durham, UK Rob Strachan, Oxford, UK John Dewey, Oxford, UK

Acknowledgements The editors thank the following colleagues and friends who kindly helped with the reviewing of the papers submitted for this volume: Ian Alsop, St Andrews, UK Arild Andreasson, Oslo, Norway Jim Andrews, Southampton, UK Rob Butler, Leeds, UK Peter Cobbold, Rennes, France Sandy Cruden, Toronto, Canada Dickson Cunningham, Leicester, UK Mike Curtis, BAS, UK Richard D'Lemos, Oxford Brookes, UK Roy Dokka, Louisiana, USA Tim Dooley, London, UK Bill Fitches, Aberystwyth, UK Mary Ford, ETH-Zurich, Switzerland Haakon Fossen, Bergen, Norway Rod Gayer, Cardiff, UK Laurel Goodwin, New Mexico, USA Reinhard Greiling, Heidelberg, Germany John Grocott, Kingston-on-Thames, UK Bob Hatcher, Tennessee, USA Becky Jamieson, Dalhousie, Canada Richard Jones, Halden Norway Steve Knott, Aberdeen, UK Peter Koons, Otago, New Zealand Shoufa Lin, GSC Ottawa, Canada

Geoff Manby, Greenwich, UK Ken McCaffrey, Kingston-on-Thames, UK Andy McCaig, Leeds, UK Tim Needham, Spilsby, UK Franz Neubauer, Salzburg, Austria Clyde Northrup, Boston, USA Terry Pavlis, New Orleans, USA John Ridley, ETH-Zurich, Switzerland Jurriaan Rjeiis, London, UK Gerald Roberts, London, UK Paul Ryan, Galway, Eire Bryan Storey, BAS, UK Brian Sturt, Trondheim, Norway Art Sylvester, California, USA Enrico Tavarnelli, Potenza, Italy Christian Teyssier, Minnesota, USA Basil Tikoff, Minnesota, USA Andrew Tomlinson, Santiago, Chile Pete Treloar, Kingston-on-Thames, UK Mike Watkeys, Durban, South Africa Alastair Welbon, Statoil, Norway John Whalley, Portsmouth, UK Nigel Woodcock, Cambridge, UK

Transpression and transtension zones J. F. D E W E Y 1, R. E. H O L D S W O R T H

2 & R. A . S T R A C H A N 3

1Department of Earth Sciences, University of Oxford, Parks Road, Oxford OX1 3PR, UK 2Department of Geological Sciences, University of Durham, Durham DH1 3LE, UK 3Geology and Cartography Division, School of Construction and Earth Sciences, Oxford Brookes University, Gypsy Lane, Headington, Oxford OX3 0BP, UK Abstract: Transpression and transtension are strike-slip deformations that deviate from simple shear because of a component of, respectively, shortening or extension orthogonal to the deformation zone. These three-dimensional non-coaxial strains develop principally in response to obliquely convergent or divergent relative motions across plate boundary and other crustal deformation zones at various scales. The basic constant-volume strain model with a vertical stretch can be modified to allow for volume change, lateral stretch, an oblique simple shear component, heterogeneous strain and steady-state transpression and transtension. The more sophisticated triclinic models may be more realistic but their mathematical complexity may limit their general application when interpreting geological examples. Most transpression zones generate flattening (k < 1) and transtension zones constrictional (k > 1) finite strains, although exceptions can occur in certain situations. Relative plate motion vectors, instantaneous strain (or stress) axes and finite strain axes are all oblique to one another in transpression and transtension zones. Kinematic partitioning of non-coaxial strike-slip and coaxial strains appears to be a characteristic feature of many such zones, especially where the far-field (plate) displacement direction is markedly oblique ( largest infinitesimal stretching direction; kl, maximum finite stretch direction. It should be noted that the ISAs have a constant orientation during steady-state deformations, whereas the finite strain axes rotate (for Wk>O) and may exchange positions. Wk = 1: simple shear (top). Wk = O: coaxial deformation (bottom). ISA2 becomes vertical (e.g. simple shear dominated). Similarly, for Sanderson & Marachini's model of transpression (type B), the orientation of ISA1 is in the horizontal plane only for 1>Wk>0.81 (simple shear dominated), and vertical for Wk 0.87 (type A) or 0.81 (type B; Fig. 4). Similarly, k2 changes to k3 for the corresponding transtensional reference deformations (types A and B). The reason for such switches is that coaxial deformations accumulate strain faster than simple shear (Tikoff & Fossen 1995). Consequently, stretching caused by the simple shear component may dominate the strain in the early history, but eventually is overcome by the increasingly effective coaxial deformation at a later stage. Switches in strain axes were first described for type B transpression by Sanderson & Marchini (1984), and occur as the strain exhibits a perfect flattening (transpression) or constriction (transtension) geometry. This coincides with the strain path touching the horizontal axis of the Flinn diagram (see A and B transpression in Fig. 5). Furthermore, the boundaries between deformations where finite strain axes do and do not switch positions during progressive steady-state deformation trace deformations where the first strain increment produces a perfect flattening geometry (constrictional geometry for transtension). In general, transpressional (transtensional) deformations in the vicinity of these lines develop strains very close to pure flattening (constriction). Fig. 5. Steady-state (constant Wk and e0 strain path in Flinn space for the five reference transpressional deformations A-E. (Note the large variations in paths and strain geometry. See text for discussion.)

TRANSPRESSION AND TRANSTENSION MODELS

21

Fig. 6. Variation in shape of strain ellipsoid for transpression or transtension. Perfect coaxial strains along bottom (Wk = 0) line. Simple shear along upper, horizontal boundary. The strain ellipsoids at the bottom apply for Wk = 0 and illustrate the shape and orientation of the coaxial components of three of the reference deformations (A, C, and E) of transpression and transtension.

More important than switching strain axes is perhaps the geometry of finite strain and the strain path in general for the different types of transpression. Figure 5 shows a selection of steady-state (constant Wk) transpression strain paths (Wk = 0.98, 0.95, 0.85, 0.5, and 0.2) (the corresponding paths for transtension are found by reflecting the transpression paths about the Wk = 1 line). The simplest pattern is developed during type D deformations, which is a plane strain sub-simple shearing. It is well known that plane strain deformations develop along the line with slope of one (Flinn k value of one), regardless of the degree of non-coaxiality of the deformation. Type C deformations develop a spectrum of strain paths between simple shear (plane strain; Wk = 1) and perfect flattening (Wk = 0). However, strain becomes increasingly oblate during deformation, and strong flattening fabrics are characteristic even for relatively low strains (unless deformation is very close to simple shear). Type E deformations behave identically to type C, except that strains are constrictional instead of flattening. Type B transpression develops flattening strains, particularly for Wk values between c. 0.6 and 0.95. Type A transpression is different because (1) any strain geometry (oblate or prolate) can be achieved (depends only on Wk), and (2) steady-state paths with Wk>0.5 may

cross the diagonal Wk = 1 line. If so, strain develops from a flattening fabric to a constrictional one for transpression (but never from constriction to flattening) and vice versa for transtension. The geometry of the strain ellipsoid is qualitatively shown in Fig. 6 for the spectrum of deformations discussed in this work. It should be noted how type D transpression and transtension separate deformations that produce oblate and prolate strain shapes, and how perfect constriction and flattening throughout deformation only occur along the base of the diagram (Wk = 0). Finite strain has implications for fabric development in deformed rocks. It is generally true that constrictional strain leads to strongly lineated rocks (L-fabrics) whereas flattening strain results in strongly foliated rocks without a lineation (S-fabrics). Intermediate cases give rise to LS- and SL-fabrics. Although it is shown above that the geometry of finite strain changes during deformation in most cases, it is possible to predict the finite fabric in rocks deformed by the various types of transpression and transtension discussed in this work. Such fabric variations are shown in Fig. 7 for reasonable amounts of finite strain. In this figure, which is based on Figs 4-6, type D transpression and transtension separate fields of S- and L-dominated fabrics, and the upper Wk=l (simple shear) boundary

22

H. FOSSEN & B. TIKOFF

Fig. 7. Variation in type of fabric for the different types of transpression and transtension discussed in the text. marks fabrics where planar and linear fabrics are equally well represented in the ideal case.

Rotation of lines Analytical work has revealed that lines and planes rotate away from a certain direction (usually a line orientation, less commonly a plane), commonly referred to as the source, fabric repellor or repulsor. Lines and planes rotate towards a different direction (or plane) known as the sink or fabric attractor (Passchier 1997). These directions are governed by the orientations of flow apophyses (see above), which are directions of maximum, intermediate and minimum rate of particle movement of the flow. Passive rotation of linear and planar markers can be modelled mathematically by use of the deformation matrix, as described by Flinn (1979) and explored for type B transpression or transtension by Fossen et al. (1994). The result of such modelling (Fig. 8) shows a range of patterns that characterize the different classes of transpression. For type A transpression, a rotational movement of line markers towards vertical is characteristic. This 'down-the-drain' pattern is only perfectly developed if the two coaxial components along the shear plane are of equal magnitude. If not, an oblique flow apophysis emerges and modifies this pattern somewhat. However, for all transpressional deformations between types A and C (Fig. 8), lines will always rotate towards the vertical direction. This pattern changes across type C transpression, where the two coaxial components along the shear plane are of equal magnitude. Here material lines rotate from an oblique, horizontal orientation along great circles until they meet the vertical shear plane and essentially stop

rotating (fly-paper effect). Hence, line orientations are expected to be distributed along the shear plane for high strains. However, if the vertical stretching is reduced or reversed in magnitude (including types D and E transpression), lines are finally parallel to the horizontal shear direction. Type E transpression is peculiar insofar as lines rotate away from an oblique plane defined by two of the flow apophyses (planar source). In general, lines are finally vertical in transpression if the vertical stretch is significant, but rotate towards the horizontal shear direction if the vertical stretch is small or negative. For transtension, on the other hand, lines tend to rotate towards a horizontal direction which is governed by the oblique flow apophysis. The exception to this general rule occurs in the case when an additional vertical component of stretching occurs (type E in Fig. 8). In this case lines are attracted to, and eventually distributed along, a plane containing the oblique flow apophysis and the z-axis (fly-paper effect). An important difference between transpression and transtension is that whereas lines rotate towards parallelism with the x- or z-axis in transpression, lines generally rotate towards a horizontal direction oblique to the shear direction (x) in transtension (compare type B, C and D transpression with transtension in Fig. 8). In both cases a large angle is expected between rotated linear structures and the shear direction.

Rotation of planes Rotation patterns for passive planar markers (poles to planes) in transpression are also shown in Fig. 8. The shortening perpendicular to the shear zone generally causes planes to rotate towards a vertical position, eventually parallel to

TRANSPRESSION AND TRANSTENSION MODELS LINES

x

PLANES

x

fU

X

x

x

@

I LINES

P

23

x

x

x

x

x

Y

Fig. 8. Rotation patterns of lines and poles to planes for Wk = 0.75. Squares are sources; triangles are sinks or attractors. Planar sources and sinks are shown by stippled lines. (See text for discussion.) the shear zone. A n exception to this is (rare) type E transpressional zones, which shorten vertically and extend in the shear direction. In this case planes end up with a common strike direction (parallel to the shear direction) but with any amount of dip. In transtension (Fig. 8, lower part), planes rotate towards a horizontal position if the shear zone perpendicular extension is mainly compensated by vertical shortening (left of type C). If shortening in the shear direction is significant (right of type C), planes rather rotate into a vertical orientation which is oblique to the shear plane.

Complicating factors The characteristics explored above for various types of transpression (and transtension), such

as finite strain, rotation patterns and fabrics, are not unique to a certain type of deformation. However, a combination of several characteristics may help to constrain the boundary conditions and kinematic vorticity. Table 1 shows the characteristic features for type A - E transpression. Although natural deformation zones may have additional complications that may restrict the usage of this table, it is useful for a general classification of the deformation in question. C o m m o n complications include heterogeneous deformation, non-steady-state deformation, and strain partitioning (see below). Strain is not likely to be h o m o g e n e o u s throughout the zone, but rather: (1) increases in intensity towards the central part of the zone, and/or (2) preferentially partitions a component of the simple shear deformation into a high

24

H. FOSSEN & B. TIKOFF

Fig. 9. Heterogeneous strain across a transpression zone can be modelled by use of several deformation boxes with different finite strain. The example shown models heterogeneous, steady-state type B transpression with maximum strain in the central part of the zone. strain zone. In the first case, the kinematic vorticity number and boundary conditions may be constant. The deformation zone can be modelled by using multiple deforming elements instead of just one (Fig. 9), and Table 1 is still applicable. However, the shape and the orientation of the finite strain ellipsoid generally change during deformation, even if W k and the angle of convergence do not. Of particular interest are the cases where k2 and ~k1 switch positions during progressive deformation (e.g. type A and B transpression; see Fig. 4). In the extreme case (type A transpression with W k around 0.6), oblate marginal strain ellipsoids may change to prolate ellipsoids towards the centre. Similarly, low-strain parts of the deformation zone may show horizontal lineations whereas more intensely deformed parts exhibit vertical lineations (switch of k2 and kl) (Fig. 4). For transtension, low-strain zones may have vertical foliation and high-strain zones have horizontal foliation (switch of )t2 and k3). Any partitioning

of the simple shear component will tend to accentuate such effects (Tikoff & Teyssier 1994; Tikoff & Greene 1997). Non-steady-state transpressional deformations are easily modelled as a number of incremental deformations with different kinematic vorticity number and/or different boundary conditions (e.g. a change from type B to type C transpression). However, the problem requires knowledge of the change in external boundary conditions, which is difficult or impossible in most cases (see, however, Fossen & Tikoff 1997). A change in convergence angle along an obliquely convergent plate margin is the simplest case that could be modelled very easily (as sequential deformations). Evidence of strain partitioning is easily observed in the field as domains of low kinematic vorticity numbers (strong component of shortening across the shear zone) between strike-slip dominated faults or narrow shear zones of high shear strain (e.g. Tikoff & Teyssier 1994; Jones & Tanner 1995; Kirkwood 1995; Goodwin & Williams 1996; Northrup & Burchfie11996).

Tectonic application of transpression and transtension Obliquely convergent plate boundaries are generally characterized by a contractional orogenic wedge tens to hundreds of kilometres wide (e.g. Vauchez & Nicolas 1991). Relative plate motion and the plate boundary are proposed to provide the primarily boundary conditions for continental deformation (e.g. McKenzie & Jackson 1983; Molnar 1992; Teyssier et al. 1995). Proof of this relation is the generally good correlation between neotectonic strain rates from geodetic surveys and relative plate motions. This relation is relatively straightforward in continental interiors and orthogonal collisional belts,

Table 1. Characteristics for the five transpressional reference deformations in Fig. 2; a similar scheme can be made for transtension, using the results" presented in the other figures and in the text Type

Strain

A transpr,

any

B transpr,

oblate

C transpr, D transpr, E transpr,

oblate plane prolate

kl (lineation) vertical (horizontal) vertical (horizontal) horizontal horizontal horizontal

Fabric

ISA 1

L, S

vertical, horizontal vertical, horizontal horizontal horizontal horizontal

LS, S S, SL S=L SL, L

Linear markers

Planar markers

--)vertical

--~vertical (any strike)

~vertical

~vertical (Hshear plane)

---)yz-plane --)horizontal ~horizontal

---~vertical(][shear plane) -~vertical 0[shear plane) any dip, strike normal to oblique flow apophysis

TRANSPRESSION AND TRANSTENSION MODELS

25

but also applies in more complex areas of oblique convergence (e.g. the San Andreas fault system; see Teyssier & Tikoff, this volume). Thus, we investigate the effect of the boundaries of obliquely convergent systems on the style of transpression. Movement

a n d slip rates

D e f o r m a t i o n within an obliquely collisional zone depends primarily on the direction of the plate motion relative to the plate margin orientation. Plate motion is commonly decomposed into a normal and tangential component (e.g. Engebretson et al. 1985). However, the relation between the plate motion and the deformation is complex, except for the case where the movement is perfectly orthogonal to the plate boundary. In regions of oblique convergence or divergence, the motion, the infinitesimal strain axes (ISAs), and the finite strain axes are not parallel (Fig. 10b). In all models of transpression (types A - E ) with any wrench component, there is an angular difference between the plate motion (oblique flow apophysis) and the fastest horizontal shortening direction (ISAHmin, equal to ISA 3 if ISA 3 is horizontal or ISA2 if ISA 3 is vertical). The finite strain long axis is originally parallel to the fastest horizontal stretching direction (ISAHmax, equal to ISA1 or ISA2), but rotates toward a margin-parallel orientation. Progressive simple shearing (wrenching) is a good end-member example of this behaviour. Experimental deformation, even applied to fracturing, for example, by Withjack & Jamison (1986), corroborates the numerical modelling. In their transtensional model (type B), normal faults form slip parallel to the ISAHmax, not the plate motion. It is a relatively common misconception in the geological and geophysical literature that the slip direction on faults (representing the infinitesimal contraction direction or ISA) is parallel to the plate motion vector (e.g. Mount & Suppe 1992; Yu et al. 1993) (Fig. 10a). This idea of 'uniaxial' deformation is not applicable to obliquely convergent boundaries, because it fails to account for the wrench component of deformation. Parallelism of the plate motion and the slip vectors only applies if (1) the deforming zone is orthogonal to the motion (oL -- 90 ~ or (2) the deforming zone is offset along a strike-slip fault so that the margin locally is orthogonal to the motion vector (Fig. 10c). Typically, the orientation of the slip vectors is at a higher angle to the plate boundary than the plate motion vector (e.g. Yu et al. 1993) (Fig. 10b), similar to

v

v

v

v

9

L . 4 ~ ~ Y Y ~ S,/~HSi AHmax b.

v',~--vPlatemotion

c

d.

\

%

x

Fig. 10. The relationship between plate motion (convergence vector), plate boundary orientation, infinitesimal strain axes (ISAHmaxand ISAHrnin), and finite strain axes in the horizontal section for a transpressional system. (a) The common misconception is that the plate motion vector is parallel to the infinitesimal contraction direction (ISAHmin) or the finite contraction direction, a geometry that cannot accommodate the wrench component. In general, the plate motion, ISAs, and finite strain axes are all differently oriented (b) except for special cases where the margin is locally orthogonal to the motion vector (c). (d) Strike-slip partitioning increases the misfit between the plate motion and the ISAHmin.

26

H. FOSSEN & B. TIKOFF

numerical predictions for transpression. For example, on the South Island of New Zealand, an obliquely convergent setting which does not exhibit strike-slip partitioning, the slip on increasingly recent faults is oriented at a systematically higher angle than that predicted from the plate motion direction (Norris et al. 1990; Teyssier et al. 1995). If strike-slip partitioning occurs, the angle of contractional slip vectors should be oriented at an even higher angle to the plate margin (Fig. 10d). Conditions leading to type B transpression or transtension The orientation of the relative plate motion, its along-strike variation, and the geometry of the plate margin will have major effects on the resultant deformation (e.g. Beck 1991; McCaffrey 1992). In a straight plate margin with a constant angle of plate convergence, type B transpression is the most likely deformation. The reasoning is based on the nature of the flow apophyses and particle paths discussed above. The map-view flow lines created in such a tectonic setting must move directly toward each other, as no gradient of deformation will occur along the collisional boundary. Rather, deformation will accumulate by vertical movement of material. In type B transpression, flow lines are straight when viewed in the horizontal plane (Fig. 2). However, vertical movement is accommodated by a pure-shear component of deformation. It should be noted that, at high crustal levels, this 'bulk' pure shear component of deformation is attained along opposite dipping thrust faults and related folds, such as in a flower structure. Conditions leading to type A , (7, D a n d E transpression or transtension A component of arc parallel stretching (Fig. 11) occurs if: (1) the plate margin has a convex orientation or (2) the angle of relative plate motion changes along strike, such that the tangential component of the relative plate motion increases along the plate margin. This situation results if the oceanic plate has a nearby pole of rotation, located within the oceanic plate (Av6 Lallemant & Guth 1990; Beck 1991; McCaffrey 1992,1996). In these cases, the material is forced to stretch parallel to the margin, to maintain compatibility. The resulting types of deformation are, with an increasing component of arc-parallel stretching, types C, D, and E. The amount of divergence and curvature will ultimately determine which deformation occurs. In

Fig. 11. The effect of divergent or convergent displacement fields and non-planar boundaries (margins) of transpression zones on the resulting type of transpression. Type B transpression would be expected from a setting with straight margins and a homogeneous displacement field (not shown). Based on McCaffrey (1992). terms of movement in the horizontal plane, curved particle paths are necessary to move material away from the colliding unit. An extreme case of margin-parallel extension occurs as a collision, either of a crustal salient or a triple-junction. The effects of such a plate margin will probably be most extreme in the fore-arc region, in which the stretching will be maximum (Av6 Lallemant & Guth 1990). Very prominent, margin-perpendicular normal faults are expected and may involve structures similar to core-complexes. Rapid uplift of deep-seated material, such as blueschists, has been proposed, as a result of such kinematics (Av6 Lallement & Guth 1990). If type D transpression is suggested, the plane-strain aspect of this deformation is easily accommodated by conjugate strike-slip faults (Weijermars 1993). The reverse case, involving a component of arc-parallel stretching, occurs if (1) the plate margin has a concave orientation, such as a

TRANSPRESSION AND TRANSTENSION MODELS large-scale salient or (2) the tangential component of the relative plate motion decreases along the plate margin, caused by a nearby pole of rotation within the adjacent continental plate (Fig. 11; Beck 1991; McCaffrey 1992, 1996). In these cases, a transition from type B to type A transpression is expected, and all of the normal component of oblique plate motion and a percentage of the transcurrent motion (resulting in orogen-parallel shortening) is accommodated by vertical uplift. Thus, this type of plate interaction is extremely efficient at uplifting rocks quickly. In the brittle regime, the margin-parallel and margin-normal shortening will both result in contractional structures. Thus, dome and basin structures may result from folding, or two separate, sub-perpendicular orientations of thrust faulting may occur. Strike-slip p a r t i t i o n i n g transpression

Fitch (1972) originally observed that the oblique plate motion is commonly accommodated by a large strike-slip fault (e.g. Great Sumatra fault) and regions undergoing primarily contractional deformation. However, as was routinely noted (e.g. Oldow et al. 1989), regions on both sides of these major structures commonly contain a large component of wrench deformation. In particular, Jamison (1991) demonstrated that a significant percentage of wrench motion is accommodated within the structures adjacent to major strike-slip faults. This process of strikeslip partitioning has been kinematically modelled for type B transpression (Fossen et al. 1994; Tikoff & Teyssier 1994; Jones & Tanner 1995; Krantz 1995) and type B transtension (Fossen et al. 1994; Teyssier et al. 1995). Any of the models of transpression or transtension (A-E) could exhibit strike-slip partitioning behaviours if the partitioning is caused by a tendency to localize deformation. However, in a dynamic explanation of the process, Tikoff & Teyssier (1994) suggested that strike-slip partitioning is caused by a major misorientation of infinitesimal strain and finite strain axes. If this reasoning is correct, strike-slip partitioning is limited to types A and B transpression (e.g. Tikoff & Teyssier 1994).

Transpressional terrane tectonics A common observation in orogenic belts is the accumulation of small fragments of oceanic or continental material, or tectonic terranes, within areas of oblique convergence (e.g. Coney et al. 1980). The western edge of North America, for instance, is a well-known example

27

of terrane accretion. Because these terranes collide obliquely with the irregular margin of North America, the kinematic models of transpression are relevant in describing their deformation. This analysis follows the concept of transpressional terranes proposed by Oldow et al. (1989). Terrane b o u n d i n g faults a n d shear z o n e s

It is commonly considered that major transcurrent motion only occurs on vertical faults or shear zones with sub-horizontal lineations. It has been repeatedly pointed out (Fossen & Tikoff 1993; Fossen et al. 1994; Robin & Cruden 1994; Tikoff & Teyssier 1994) and documented in geological field examples (Hudleston et al. 1988; Tikoff & Greene 1997) that the lineation direction in transpressional shear zones does not reflect the tectonic movement direction of the simple shear component of deformation. Strain modelling indicates that it takes a relatively low angle of convergence (c~ .......... -.-" , - - ' "................. ..". . . . § . +. . . ~ . . + " .... ~ 1--. ~,._.' , "F t 4 ~ ................. ~ 4zs"/] + ~, ~ ./ ..... ~ -/ ~ 71 ./r ..... / ! ~r ' '/ + + 5b 48 ~/" ~ " ! ~ ! ! + 4+ + +

!

'

,', 4-

,

,

I I

+

Centraldomain

!

2. ! {

,.'

4,

!

/,~4 Lu-. b, ~37 + +

.

: , , / \ ' , ~

+

'

+

.

+/ + + ! " ! / "" ~ 6 4

,,!

,"

"

:/ ~

46~

+

, ";

,',

•T

.

4-

/-i-,

t

..'+ !

+

/ -T

.

,': ,.r

."

! {

! +

;

t

! +

/

~

Z ~

, +

Z60 L\ ~'

!

~

//65

!

+

, 4-

57 . . . . , , I ! r + + + + / ! J ! / . . . . . .

4~

~

, § .... _r

+

!

~-

-

,

+

,

+

z .

.

764 ~

+

!

,

+

/

+ !

+

,

I

+

,

)

+." -,

~

+

.

.z60 L~.

/

+ +

+

t .

.

+

',

.

-

+

+

. . §. .+ . ~.

+

+

,

+

. ___ ~u3 ~

+ !

!

4-

!

{

!

r e,. ::.T

+

.

4-

,

+ !

+

// / t i , , , t , %," ~ + + + + + + + +

!

/ {

/

,

+ /

+

+'-/+

t

~

~

t

,

+ /

,'+'+'+'+Marginaldomain

+

+

,% ..........

+

!

+

+

.

+

~

+

+

+

+,"

,,~'~'+,4-,+'+,4-,

? '

+' +, /

r

1

!

+

+

+

+/+

.

+

!

+ /

+

+ .'! / / I / / ! / ! ! / + .'+ + + + + + + + + + + +

+

+

52 +

/ +

~

+ /

+

~;"

.

.

+

/

+ /

+

+

+

+

,

+"! z ~ / :" ' ~=7 . . I + + + + + " I ,Ju.J ~ /-64 /.~' + ~ + + + + ,/+ + '' , ',+, ,,~t- L \ ..... ~ + + + .t t f / - " 9 k~ ! / / .... + + + 4 4~ 4; .... : '=8"" + " + + + + + ,r" ~ ~ ! ! ' ! : ' + + 4- + + + + + .~.,'.:; + + + + + . . .... . :. " . . . . +

I

+

+ +

+

"

~

+

. . . . .

+ 4-

+

+

.. //../+ .I "

~"

+.,

]"

+ !

+

+

+

/

/ .

+ t

§

! ~

!

+ + + .,~-

+

+.,;

§ ,."

+" "+

+

+

+

+

+

.."

+.

.

,

t

§

t/.+

.# =~ + ,~.oo +..; ~ "~

+

.,

. . . . . bmauononu

+

++ . . . . +

1-

+

~+

+

+

+

+

+

-t+

+

+

1-

+

+

1+

+

+

1-

-f

+

-t+

+

+

rock

t+

+

deformed

Country

+ +

4

43

/

7.82'

~'

r

2.76 (or Wk > 0.81), it is horizontal when the strain is low and may 'switch' to down-dip when the finite strain increases to a critical value (Fig. 9a; see Sanderson & Marchini (1984) and Fossen & Tikoff (1993) for discussion). In transpressional zones with 0~ < + < 90~ the lineations plot away from the vorticity-normal section (VNS) and close to the dip line of the zone, where the values of ~//~are low (this, for a general transcurrent zone, means steeper lineations, because the dip line of a transcurrent zone is nearly vertical). The lineations approach the VNS as "9/~increases (which means shallower lineations for a transcurrent zone). When the value of -~/~is high (>10) and for low to intermediate finite strain magnitudes (hal/z/h31/2 less than c.100), the locus of the lineations is virtually on the VNS (Fig. 9a), a feature to be expected of a monoclinic shear zone. Therefore a monoclinic structural geometry can result from a triclinic movement picture with a large boundary-parallel movement component. As finite strain accumulates, the lineations on all "9/~ paths and for various + values converge towards the dip line of

50

S. LIN E T A L .

k3

~6

k.-..~-;-----~...'il .....

77

133

,=40 ~

X2

~

1

X3

~N~~.~,,qt4-S

J,~2

~ . ,~,,~

-~

rdire

x~

(a) Transpression 133

nn

I x~ ~--- V N S ------r~\.,,--

(~ ....

I

~, I

)

:_~

>~

v . ~ .... :,_~ >1 Xs

TRANSPRESSION ZONES OF TRICLINIC SYMMETRY

51

(b) Transtension

.

oO:,,,~ ~,~__/n/ ,"Z'" ~~"~-~% % I~~"se~dNct,o

.

shear direct/ion

Fig. 9. Equal-area lower-hemisphere projection of variation and evolution with time of finite-strain-related lineations (ha) and poles to finite-strain-related foliations ()~3) in a transpressional (a) or transtensional (b) zone with varying ~/~ and +. Numerals 1, 2, 4, 6 and 20 are values of ~/~. White arrows indicate the evolution with time (increasing finite strain) for various ~/~ values. The dashed white arrow on the diagram of qb = 0~ indicates 'switch' of orientation of stretching lineations from horizontal to vertical. VNS, Vorticity-normal section; SZB, shear zone boundary, xl, x2, x3 correspond to those in Fig. 6. Let us take a transpressional zone with + = 20 ~ as an example. When ~/~ is low (e.g. < 3), the lineation is nearly down-dip. As "~/~increases, the lineation is closer to lying on the VNS. On an average foliation, the lineation across a transpressional zone swings from nearly down-dip where ~/~ is low to nearly strike-parallel where 4//~is high. As finite strain increases the lineations converge to being steeper though with significantly different rates of convergence. When ~/~ is high (>10) and for low to intermediate finite strain magnitudes (~k11/2/~.31/2less than c.100), the lineations essentially lie on the VNS, and the symmetry of the structures becomes monoclinic within the resolution of observation, although the movement picture is trielinic. The shear zone boundary is shown as vertical here. Geometry for non-vertical shear zones can be obtained by rotating these diagrams (e.g. Fig. 11). (See text for further discussion.) the shear zone. Thus, the e n d orientation of the lineation is not the shear direction of the simple shear c o m p o n e n t but the dip line of the zone (close to a vertical line for a general transcurrent

zone). For a general transcurrent zone (+ is low but not 0 ~ and the zone b o u n d a r y is n e a r vertical), the lineations swing from nearly down-dip w h e r e the ~/~ ratio is low to nearly strike-parallel w h e r e

52

S. LIN E T A L .

Fig. 10. Plots of the shapes of the finite strain ellipsoids for transpressional or transtensional zones of constant strike length for various qbvalues. Transpression produces oblate strains (K 1). Simple shear produces plane strain (K = 1). Solid square, solid circle, cross and open square represent, respectively, "~/4= 2, 4, 6 and 10.

the "~/4ratio is high on an average foliation (see the diagram of + = 20~ in Fig. 9a). The poles to foliations in the transpressional situation, however, always plot near the VNS (Fig. 9a). In contrast, in transtensional zones, the lineations plot close to the VNS, very close to it when the finite strain is low and slightly away from it as the finite strain increases. Poles to foliations in transtension are variable and similar to the lineations in transpressional situations (Fig. 9b). Figure 9a shows that stretching lineations in transpressional zones can (1) lie on or very close to the VNS and, as strain increases, approach the

shear direction for monoclinic or apparent monoclinic (~?/~> 10) situations; (2) be perpendicular to the shear direction when + = 0~ and hi is parallel to x3; (3) generally be inclined to both the VNS and the shear direction in triclinic situations. The simple statement made for simple shear zones that stretching lineations will align with, and therefore indicate, the shear direction cannot be extrapolated to three-dimensional transpressional (or transtensional, Fig. 9b) shear zones. Figure 10 presents plots on Flinn diagrams of the shapes of the finite strain ellipsoids in transpressional and transtensional zones of constant

TRANSPRESSION ZONES OF TRICLINIC SYMMETRY

53

strike length. They show the evolution with time of the shape of the finite strain ellipsoids for various values of ~/~ and dO.It is readily seen that transpression produces oblate strains (K 1).

Interpretation and discussion Application of theoretical modelling results The results of the theoretical modelling should be applied to natural shear zones with care and the following points should be kept in mind. (1) The present model assumes both isochoric deformation and constant strike length of the shear zone. A more general model that includes triclinic flow, volume change and strikelength change is given in Jiang & Williams (in press). (2) The theoretical model is for a homogeneous domain and a steady period, and the more closely these conditions are approximated by a natural shear zone the better the comparison will be. Heterogeneously deformed shear zones are best treated by dividing them into domains that better approximate the homogeneous condition, as is generally done in structural analysis, and the results of the present modelling can be readily applied to such shear zones. The heterogeneity arises in a number of ways. Not only strain magnitude, but also do and -)/~, may vary from point to point. Thus different parts of a shear zone may have different do value paths, different -9/~paths, and/or have been at different positions along any one of these paths. Structural data from natural shear zones do not necessarily, and generally do not, define complete paths as shown in Fig. 9. (3) We have assumed here that lineation and foliation are approximately parallel to the finite strain axes (kl and )tl)t2, respectively). This is not always true and the theoretical models should only be applied where there is good reason to believe that the assumption is valid. Taking the Roper Lake shear zone as an example (Fig. 4a), finite strain-related stretching lineations (Ls) were carefully differentiated from the 'ridge-ingroove'-type striations (Lc). Only Ls and poles to S-foliations are compared with the theoretical results. However, C-surfaces give an approximation of the shear zone boundary and plots of Lc indicate the shear direction (Figs 4a and 11). The scatter of Ls plots reflects the heterogeneity of the shear zone. The most significant point of these data is that the statistically defined mean for the stretching lineation of the marginal domain deviates from the great circle girdle defined by the data of the central domain.

Fig. 11. Comparison of the strain geometry of the Roper Lake shear zone (as shown in Fig. 4a) with the strain geometry predicted for a dextral oblique transpression zone with + = 50~ that strikes 45~ and dips 75~ The Roper Lake shear zone is interpreted as such a transpression zone with a higher ~/~ ratio in the central domain than in the marginal domain. Comparison of the Ls data with the evolution of ~k1 with time indicates that the central domain is more evolved (more deformed) than the marginal domain, consistent with the observations described above. The shaded areas are plots of mS and Ls from Fig. 4a. G, C, and Lc also correspond to those in Fig. 4a. Numerals 2, 4, 6 and 10 are values of "~/~..White arrows indicate the evolution with time (increasing finite strain) for the various "~/~values. VNS, Vorticity-normal section; SZB, shear zone boundary. Because lineations in a transpression zone may rotate as passive lines towards being steeper after their formation (Fossen et al. 1994), a + value obtained by comparing the observed with the predicted geometry may only be a maximum estimate.

Interpretation of the Roper Lake shear zone The structural geometry of the R o p e r Lake shear zone is characterized by variable orientations of the lineations and relatively constant orientations of foliations. Comparison with the results of the theoretical modelling indicates that the shear zone can be interpreted as a transpression zone with oblique boundary-parallel motion (i.e. oblique transpression) and a higher ~/~ ratio in the central domain than in the marginal domain (Figs 11 and 12). The latter explains why the stretching lineation is shallower in the former domain and steeper in the

54

S. LIN E T A L . quartz c-axis fabrics are stronger, in the central domain. These all indicate localization of the simple shear component (4/) there. Thus, we suggest that variation in the 4//~ratio is due more to the localization of the simple shear (4/) than to variation in the pure shear (~) across the shear zone (Fig. 12). As discussed below, oblique transpression (0 ~ < + < 90~ (and thus a triclinic movement picture) and variation in the 4//~ratio as a result of localization of 4/ are probably common features. O b l i q u e transpression a n d triclinic m o v e m e n t picture

Fig. 12. Schematic diagrams showing a general model of transpression zones. (a) Pre- deformational geometry. (e) Geometry after transpressional deformation. In the model, boundary-normal compression (~) and boundary-parallel shearing (4/) take place simultaneously. To assist visualization, the deformation component related to ~ is shown as (b). The narrow zone in (b) indicated by the dashed lines corresponds to that in (c) in which 4/is concentrated. The main features of the model are that (1) boundaryparallel shearing (4/) is oblique and contains both strike-slip (ss) and dip-slip (ds) components, and (2) boundary-parallel shearing (4/) is much more localized than boundary-normal compression (~). The movement picture is of triclinic symmetry. It should be noted that the narrow zone in which boundaryparallel shearing (4/) is localized does not have to be in the centre of the wider zone under boundary-normal compression (~) as shown here. latter domain. As described earlier, C-surfaces, asymmetrical mica fish and an oblique shape fabric in recrystallized quartz grains are much more intensely developed or only observed, and

Observations of present plate motions suggest that most convergent plate boundaries have an oblique displacement vector (e.g. Liu et al. 1995). Such an oblique convergence is often heterogeneously accommodated as a result of slip partitioning; i.e. the net oblique displacement is partitioned into a more dip-slip c o m p o n e n t accommodated at the subduction zone and a component accommodated in the overriding and underriding plates via intraplate deformation (e.g. Fitch 1972; DeMets 1992; McCaffrey 1992; Shen-Tu et al. 1995). The intraplate d e f o r m a t i o n shows evidence of further slip partitioning (e.g. Gao & Wallace 1995). However, a complete partitioning where the net oblique slip is partitioned into two end members: pure dip-slip and pure strike-slip components, as argued by, for example, Tikoff & Teyssier (1994, and references therein), may be rare. Liu et al. (1995) defined a parameter K to measure the degree of slip partitioning for a subduction zone. K = 0 (e.g. in northeastern Japan) and K = 1 (e.g. in New Hebrides) represent zero and complete partitioning, respectively (Liu et aL 1995, fig. 6). It is probable that K generally lies between zero and one, as for example in the Aleutians where K --~ 0.34 (Liu et al. 1995, fig. 6). The general incompleteness of slip partitioning implies that the imposed boundary displacements for many deformation zones are oblique; the boundary displacement vector may lie anywhere in the spectrum from dip-slip (+ = 90~ to strike-slip (+ = 0~ This gives rise to triclinic movement pictures for the internal deformation. Shear zones with triclinic movement pictures are probably far more common than reported in the literature. The observed monoclinic symmetry in natural deformation zones could well be reflecting a high 4//~ratio rather than a true monoclinic m o v e m e n t picture. As d e m o n s t r a t e d above, when the ratio of 4//~ is high, the structures and fabrics in an oblique transpressional zone (having a triclinic movement picture) will

TRANSPRESSION ZONES OF TRICLINIC SYMMETRY

55

Fig. 13. Schematic diagram showing the distribution, localization and partitioning of deformation across a convergent plate boundary. The boundary-normal motion results in crustal thickening (~) which is distributed over a wide region. The boundary-parallel motion resulting in shearing (9) is often localized into narrow deformation zones or faults. These zones can be transpressional if there is an appreciable ~ component, bulk simple shear if the ~ component is negligible relative to 9, or discrete faults where the deformation is extremely localized. It should be noted that these zones or faults may have the same or different degrees of movement obliquity (+), depending on the plate motion obliquity, slip partitioning and distribution. Integration of the motions accommodated by all these zones gives the relative motion of plates. exhibit monoclinic symmetry within the resolution of observation. It is likely that many natural shear zones described in the literature are only the equivalents of the central domain of the Roper Lake shear zone described above, in the sense that they only represent more intensely deformed portions (with higher ratios of "9/~) of much wider shear zones. W i t h o u t the less intensely deformed portions (or the equivalents of the marginal domain of the Roper Lake shear zone) being considered together with the more deformed portions, the potential triclinicity of these shear zones cannot be easily recognized, as exemplified by the Roper Lake shear zone.

Localization o f simple shear Studies of ancient shear zones (including the Roper Lake shear zone described above) and observations of current plate-boundary deformation indicate that in shear zones and orogens alike, the c o m p o n e n t of boundary-parallel motion (simple shear component -~) tends to be localized whereas the component of boundarynormal motion (pure shear component ~) tends to be widely distributed (see Gordon (1995) and references therein). Shear zones are localized features of much wider orogenic belts (Fig. 13). Theoretical models lead to the same conclusion. For example, for power law rheology with stress

exponents n = 3 and n = 10, the length/width ratio for a pure compression or extension zone is one and two, respectively, whereas it is five and ten, respectively, for a simple shear zone (England et al. 1985; England & Jackson 1989; Sonder & England 1986). This means that boundary-parallel motions are approximately five times more localized than boundary-normal motions. There are other factors that may contribute to this phenomenon. (1) Boundary-normal motion results in thickening or thinning of the crust leading to buoyancy forces that make further thickening or thinning more difficult (see also England et al. 1985; England & Jackson 1989). (2) Simple shear is believed to be a fabric weakening process whereas pure shear is believed to be a fabric hardening process (Williams & Price 1990; P. F. Williams unpublished). Much of deformation at the granular scale tends to take place along weak grain boundaries. Simple shear aligns elongate minerals and therefore makes grain boundaries ever closer to the shear plane orientation, thus making grain boundary sliding (diffusion dependent or frictional) an increasingly effective deformation mechanism. Pure shear, on the other hand, aligns grain boundaries progressively more perpendicular to the shortening direction. Thus grain boundary sliding becomes increasingly

S. LIN E T A L .

56

difficult a n d f u r t h e r d e f o r m a t i o n r e q u i r e s s t r o n g e r i n t r a g r a n u l a r m e c h a n i s m s to o p e r a t e . We t h e r e f o r e c o n c l u d e t h a t l o c a l i z a t i o n of b o u n d a r y - p a r a l l e l m o t i o n w i t h i n a w i d e r z o n e of c o m p r e s s i o n or e x t e n s i o n p e r p e n d i c u l a r to t h e z o n e b o u n d a r y is a g e n e r a l p h e n o m e n o n . L o c a l i z a t i o n d o e s n o t r e q u i r e p r e - e x i s t i n g surfaces of w e a k n e s s (e.g. faults, s h e a r zones, lithological b o u n d a r i e s , weak layers and/or r h e o l o g i c a l a n i s o t r o p y ) (cf. J o n e s & T a n n e r 1995), a l t h o u g h t h e p r e s e n c e of s u c h surfaces or zones will facilitate t h e process. The work is supported by the Geological Survey of Canada through funding to S. Lin and a contract to D. Jiang from the Canada-Nova Scotia Co-operation Agreement on Mineral Development (1992-1995), and by the Natural Sciences and Engineering Research Council of Canada through a research grant to P. F. Williams and a Post Doctoral Fellowship to D. Jiang. The manuscript was improved by reviews from H. Fossen, L. B. Goodwin, S. Hanmer and S. B. Lucas. This paper is Geological Survey of Canada Contribution 1996467.

R

e

f

e

r

e

n

c

e

s

CARON,A. & WILLIAMS,P. E 1988. The kinematic indicators of the Love Cove Group in the northeastern Newfoundland. Geological Association of Canada Program with Abstracts, 13, A17. DEMETS, C. 1992. Oblique convergence and deformation along the Kuril and Japan trenches. Journal of Geophysical Research, 97, 17615-17625. DUNNING, G. R., BARR, S. M., RAESIDE, R. P. & JAMIESON, R. A. 1990. U - P b zircon, titanite, and monazite ages in the Bras d'Or and Aspy terranes of Cape Breton Island, Nova Scotia: implications for igneous and metamorphic history. Geological Society of America Bulletin, 102, 322-330. ENGLAND,P. C. & JACKSON,J. 1989. Active deformation of the continents. Annual Review of Earth and Planetary Sciences, 17, 197-226. - - , HOUSEMAN, G. A. & SONDER, L. J. 1985. Length scales for continental deformation in convergent, divergent, and strike-slip environments: analytical and approximate solutions for a thin viscous sheet model. Journal of Geophysical Research, 90, 3551-3557. FITCH, T. J. 1972. Plate convergence, transcurrent faults, and internal deformation adjacent to Southeast Asia and the western Pacific. Journal of Geophysical Research, 77, 4432--4460. FOSSEN, H. & TIKOFF,B. 1993. The deformation matrix for simultaneous simple shearing, pure shearing and volume change, and its application to transpression-transtension tectonics. Journal of Structural Geology, 15, 413-422. , & TEYSSIER, C. 1994. Strain modeling of transpressional and transtensional deformation. Norsk Geologisk Tidsskrift, 74, 134-145.

GAO, L. & WALLACE, Y. C. 1995. The 1990 Rudbar-Tarom Iranian earthquake sequence: evidence for slip partitioning. Journal of Geophysical Research, 100, 15317-15332. GOODWIN, L. B. & WILLIAMS,P. E 1996. Deformation path partitioning within a transpressive shear zone, Marble Cove, Newfoundland. Journal of Structural Geology, 18, 975-990. GORDON, R. G. 1995. Plate motions, crustal and lithospheric mobility, and paleomagnetism: prospective viewpoint. Journal of Geophysical Research, 100, 24367-24392. JIANG, D. & WHITE, J. C. 1995. Kinematics of rock flow and the interpretation of geological structures, with particular reference to shear zones. Journal of Structural Geology, 17, 1249-1265. JONES, R. R. & TANNER, P. W. G. 1995. Strain partitioning in transpression zones. Journal of Structural Geology, 17, 793-802. • WILLIAMS, P. F. High-strain zones: a unified model. Journal of Structural Geology, in press. KRANTZ, R. W. 1995. The transpressional strain model applied to strike-slip, oblique-convergent and oblique-divergent deformation. Journal of Structural Geology, 17, 1125-1137. KROHE, A. 1990. Local variations in quartz C-axis orientations in non-coaxial regimes and their significance for the mechanics of S-C fabrics. Journal of Structural Geology, 12, 995-1004. LIN, S. 1992. The stratigraphy and structural geology of -

-

the southeastern Cape Breton Highlands National Park and its implications for the tectonic evolution of Cape Breton Island, Nova Scotia, with emphasis on lineations in shear zones. PhD thesis, University of New Brunswick, Fredericton. 1993. Relationship between the Aspy and Bras d'Or 'terranes' in the northeastern Cape Breton Highlands, Nova Scotia. Canadian Journal of Earth Sciences, 30, 1773-1781. 1995. Structural evolution and tectonic significance of the Eastern Highlands shear zone in Cape Breton Island, the Canadian Appalachians. Canadian Journal of Earth Sciences, 32, 545-554. & WILLIAMS, R E 1992a. The geometrical relationship between the stretching lineation and the movement direction of shear zones. Journal of Structural Geology, 14, 491-497. --& -1992b. The origin of ridge-in-groove slickenside striae and associated steps in an S-C mylonite. Journal of Structural Geology, 14, 315-321. & -1993. A transpressional model for interpreting the variation in the pitch of the stretching lineation across a shear zone. Geological Associ-

-

-

-

-

-

-

ation of Canada - Mineral Association of Canada Program with Abstracts, 18, A60. LISTER, G. S. & WILLIAMS, P. E 1979. Fabric development in shear zones: theoretical controls and observed phenomena. Journal of Structural Geology, 1, 283-297. LIU, X., MCNALLY,K. C. & SHEN, Z.-K. 1995. Evidence for a role of the downgoing slab in earthquake slip partitioning at oblique subduction zones.

TRANSPRESSION ZONES OF TRICLINIC SYMMETRY

Journal of Geophysical Research, 100, 1535115372. MCCAFFREY, R. 1992. Oblique plate convergence, slip vectors, and forearc deformation. Journal of Geophysical Research, 97, 8905-8915. MEANS,W. D., HOBBS, B. E., LISTER, G. S. & WILLIAMS, E E 1980. Vorticity and non-coaxiatity in progressive deformations. Journal of Structural Geology, 2, 371-378. RAMBERG, H. 1975. Particle paths, displacement and progressive strain applicable to rocks. Tectonophysics, 28, 1-37. RAMSAY, J. G. 1980. Shear zone geometry: a review. Journal of Structural Geology, 2, 83-89. & GRAHAM,R. H. 1970. Strain variation in shear belts. Canadian Journal of Earth Sciences, 7, 786-813. RAESIDE, R. P. & BARR, S. M. 1992. Geology of the northern and eastern Cape Breton Highlands, Nova Scotia. Geological Survey of Canada Paper 89-14.

ROBIN, P.-Y. E & CRUDEN,A. R. 1994. Strain and vorticity patterns in ideally ductile transpressional zones. Journal of Structural Geology, 16, 447-466. SANDERSON, D. J. & MARCHINI,W. R. D. 1984. Transpression. Journal of Structural Geology, 6, 449-458. SCHMID, S. M. & CASEY, M. 1986. Complete fabric analysis of some commonly observed quartz Caxis patterns. In: HOBBS, B. E. & HEARD, H. C. (eds) Mineral and Rock Deformation: Laboratory Studies. Geophysical Monograph, American

57

Geophysical Union, 36, 263-286. SHEN-TU, B., HOLT, W. E. & HAINES, A. J. 1995. Intraplate deformation in the Japanese Islands: a kinematic study of intraplate deformation at a convergent plate margin. Journal of Geophysical Research, 100, 24275-24293. SONDER,L. J. & ENGLAND,P. C. 1986. Vertical averages of rheology of the continental lithosphere: relation to thin sheet parameters. Earth and Planetary Science Letters, 77, 81-90. STARKEr, J. 1977. The contouring of orientation data represented in spherical projection. Canadian Journal of Earth Sciences, 14, 268-277. TIKOVF,B. & FOSSEN,H. 1993. Simultaneous pure and simple shear: the unifying deformation matrix. Tectonophysics, 217, 267-283. - & TEVSSIER, C. 1994. Strain modeling of displacement-field partitioning in transpressional orogens. Journal of Structural Geology, 16, 1575-1588. TRUESDELL, C. A. 1953. Two measures of vorticity. Journal of Rational Mechanical Analysis, 2, 173-217. & TOUPIN,R.A. 1960. The classic field theory. In: FL~3GGE,S. (ed.) Encyclopedia of Physics, Volume III: Principles of Classical Mechanics and Field Theory. Springer-Verlag, Berlin, 226-793. WILLIAMS, P. E & PRICE, G. P. 1990. Origin of kinkbands and shear-band cleavage in shear zones: an experimental study. Journal of Structural Geology, 12, 145-164.

Analogue modelling of faulting in zones of continental transpression and transtension GUIDO

SCHREURS t & BERNARD

COLLETTA 2

1Geological Institute, University o f Bern, Baltzerstrasse 1, CH-3012 Bern, Switzerland (e-mail: schreurs@geo, unibe, ch) 2Institut Franfais du POtrole, P.O. B o x 311, F-92506 Rueil Malmaison, France Abstract: Experiments were performed to simulate deformation in zones of continental

transpression and transtension. Stratified models consisted of brittle analogue materials overlying a thin layer of viscous material. Oblique deformation was obtained by combining a basal, distributed strike-slip shear component with either transverse shortening (transpression) or transverse extension (transtension). In transpression experiments the imposed ratio of shear strain rate and shortening strain rate exerts an important control on initial fault evolution in the brittle layers of the model. In those experiments with a relatively high strain rate ratio (>3.6), subvertical, en echelon strike-slip faults develop first, striking at angles of 25-37 ~ to the shear direction. With increasing strain several convergent strike-slip fault zones form displaying positive flower structures. In low strain rate ratio experiments (_200 km long) and oblique-slip thrust faults with opposing vergence (Fig. 13). In plan and cross-section the structures of the Western Altai show strong similarities to advanced stages of our high ratio analogue models. Cunningham et al. (1996) also described the southern termination of a major dextral strikeslip fault in the Jargalant Range of the Western Altai (Fig. 14). Displacement along the fault system is accommodated by thrust faulting, oblique-slip faulting and uplift within the Jargalant Range. Cross-sectional geometry of the Range is that of an asymmetric flower structure. This geometry is similar to fault evolution from

ANALOGUE MODELLING OF FAULTS

75

Fig. 12. (a) Location of Confidence Hills within the southern Death Valley fault zone (SDVFZ) in eastern California and (b) 3D synoptic model of part of the Confidence Hills. Modified after Dooley & McClay (1996).

Fig. 13. Block diagram interpretation of part of Mongolian Western Altai, showing the High Altai, Sutai and Jargalant structural domains, each consisting of large-scale flower structures related to dextral transpressional strike-slip fault systems. Modified after Cunningham et al. (1996).

Fig. 14. Map showing location of Jargalant Range (Mongolian Western Altai). Dextral displacement along the Har Us Nuur fault system is accommodated by thrust faulting, oblique-slip faulting and uplift within the Jargalant Range (from Cunningharn et al. 1996).

76

G. SCHREURS & B. COLLETTA

Fig. 15. (a) Simplified structural map of the northwestern part of South America, showing the location of the Merida Andes and the dextral Bocono Fault in Venezuela (modified after Colletta et al. (1997). (b) Schematic crustal block diagram through the Merida Andes. Location of section is indicated in (a). stage 1 to 2 in the high ratio experiment 1661 (Fig. 3a-d), in which laterally propagating subvertical dextral strike-slip faults terminate by swinging around into oblique-slip reverse faults. A natural example of low ratio transpression experiments is possibly provided by the Merida Andes in Venezuela (Fig. 15a and b). This mountain belt is the result of Neogene dextral transpression between the Maracaibo plate and stable South America (Colletta et al. 1997). Oblique convergence is partitioned in the upper crust, with two conjugate northwest- and southeast-verging contractional structures, orthogonal to the plate boundary, and a major

intervening northeast-striking dextral strike-slip structure, the Bocono fault (Colletta et al. 1997). In the interpretation by Colletta et al., the steep Bocono fault is confined to the tectonic wedge and does not extend deeper into the crust than the sole thrust of the allochton. The structures in the transtension experiment are similar to those observed in the North Aegean Sea (Fig. 16), where dextral transtension has occurred since Late Miocene time (Lyberis 1985). Deformation occurs by a combination of en echelon arranged, E N E - W S W striking faults with normal character and subparallel striking steep faults with important

ANALOGUE MODELLING OF FAULTS

77

Fig. 16. Tectonic sketch map of North Aegean region (modified after Pavlides et al. (1990)). dextral strike-slip displacement (Pavlides et al. 1990). NW-SE striking faults have possibly originated as sinistral faults formed between major dextral fault zones and would correspond in that case to antithetic, sinistral strike-slip faults in the experiment which form after major dextral strike-slip faults have developed.

Conclusions Fault patterns at different stages of each experiment provide important information on overall kinematics, local stress field modifications, and partial partitioning of fault motion. The early fault style in transpression experiments clearly depends on the applied strain rate ratio. This ratio determines whether initial failure in the brittle layers is accommodated by steep strike-slip faults or by thrust faults. In high strain rate ratio experiments (_>3.6) steep strikeslip faults (dipping at 80-90 ~ formed early. Their en echelon arrangement can be used as an indicator of the overall sense of shear (i.e. a leftstepping pattern indicates dextral shear). The sigmoidal trace of strike-slip faults that laterally become oblique-slip reverse faults can also be used as kinematic indicator ('lazy' z-shape for dextral shear and 'lazy' s-shape for sinistral shear component). Both these kinematic indicators occur also in pure strike-slip fault systems (Naylor et al. 1986; Mandl 1988; Schreurs 1992; Richard et al. 1995). The difference with pure strike-slip fault systems is the obliquity of surface strike (with respect to the regional shear zone boundaries) of early strike-slip faults, which is larger in transpression experiments. Obliquity of surface strike in transpression

experiments increases as the strain rate ratio decreases. In low strain rate ratio experiments (100 m) vertical offsets of Cretaceous-Eocene formations in places, although existing kinematic descriptions emphasize strike-slip (Nammour 1992). Despite these throws, the majority of the faults have no landscape expression. They appear to be planed off by the oldest marine terrace on the plateau, a Miocene landscape feature (Fig. 9a; Nammour 1992). Only the Zrariy6 Fault and its relays have visible fault scarps that apparently offset the upper terrace level. We investigated this fault (Fig. 9b) where it is marked by a 30 m scarp. Its base has a concave form, the cliff-base boulders containing Lithophaga. The scarp is clearly a relict sea cliff, now over 100 m above sea level.

T h e R o u m Fault Z o n e

96

R.W.H. BUTLER ETAL.

EVOLUTION OF THE LEBANESE RESTRAINING BEND

300m

~

,.,

/ @~ 9 ~.. .

~

.

Y

.

---2_-

Nj

97

:i : "

,~

..>..

9 ir

~

7/I

~ ~

, ,.-

9

"" "

",,\

M ~ 1 7 6 1 7 6 F,.

.oo;o0 ~ .A I _

j-

,,.,\ /

Hommooa Fm

cautr zooes

beddinq

j

-

,,,,

.

.

',,,, ,

/

/

C

foulted

/ cave

mfill

( 1/

~-a~~, y

+,oo

]

/:/ b Fig. 10. Sketch map (a), cross-section (b) and structural data (c: fault planes and striae) for the Roum Fault Zone at the Damour canyon. (See Fig. 2 for location.) Selected representative fault kinematic data are illustrated as 'pseudo-focal mechanisms' (see Butler et al. 1989). Much of the D a m o u r section exposes fault zones which cut and shatter palaeokarstic features (Fig. 9c). These include large volumes of compacted red muds and speleothems, the collapsed remnants of old cave systems. These must have existed w h e n this area lay above local watertable and consequently post-date the M i o c e n e marine terrace. Thus a substantial part of the d e f o r m a t i o n in the Darnour section post-dates the late M i o c e n e transgression and subsequent

regression. We conclude that at least these portions of the R o u m Fault post-date the strike-slip on the n o r t h e r n part of the Y a m m o u n e h .

Estimating post-Messinian offsets on the R o u m Fault O u r o b s e r v a t i o n s s u p p o r t t h e c o n t e n t i o n of Girdler (1990) that the R o u m Fault is a m a j o r zone of left-lateral d i s p l a c e m e n t that has b e e n

Fig. 9. Geomorphological features in SW Lebanon. (a) Upper marine planation surface of the Tyre-Nabatiy6 plateau, looking north across the Zahrani river valley. This surface here has an elevation of 350400 m and truncates all faults. It is inferred to be of late Miocene age (Nammour 1992). (b) Cliff line on the Zrariy6 Fault, interpreted as an un-faulted palaeo-seacliff, a pediment of conglomerate with Lithophaga and minor marine sediments, inferred to be Miocene in age. (e) Detail of palaeo-karst red soil involved within the Roum Fault Zone. Minor strike-slip faults cut this feature and the palaeo-karst is shattered adjacent to fault zones.

98

R.W.H. BUTLER E T A L . 20 km

/

/

IRUT

cofchm@nf

KELB

gorgo w,nd gap

"'""'"

....

Sidon

111 O

Present

(Okra)

b

15kin displacement

1 /

fC 30kin displacement

Fig. 11. Drainage basins and their offset across the Roum Fault. (a) Modern drainage basins and gorges across the Roum Fault. The sea-bed depths are after Hall et al. (1994). Z, Zahrani catchment. Quaternary basins are stippled. (b) A restoration of displacement on the Roum Fault which places the headwaters of the Zahrani river above the lower Litani canyon. In Pliocene and earlier times the Litani river flowed through what are now windgaps into the Hula valley. This restoration is not favoured as it offers no explanation of windgaps at the head of the modem Debb6 river. (e) The preferred restoration, which matches the headwaters of the Aouali river with the lower Litani canyon. The ancestral Zahrani headwaters may have fed the Debb6 valley.

active since the Miocene. Like the Yammouneh Fault, it is not a single fault but rather is represented by relays and strands which define a broad damage zone. Despite its seismogenic character, the Roum Fault is generally inferred to have accumulated little displacement and therefore to be a very young structure. Walley (1988) estimated displacements using river offsets, recording 9 km displacement of the Litani and 3.5 km of the Zahrani, and no mapscale offsets on the Aouali and Damour rivers. These offsets imply a dramatic displacement gradient on the Roum Fault Zone but are probably an artefact of the geomorphological analysis. The following discussion (Fig. 11) briefly reviews the results of a much larger geomorphological analysis that will be the subject of a forthcoming paper. The watershed running parallel to the R o u m Fault Z o n e between the Zahrani and Litani rivers contains a series of windgaps which betray the process of stream capture. Clearly, the modern courses of the rivers only record part of the displacement history of the fault. To establish a longer-term picture of fault displacement we now analyse the pattern of drainage basins,

windgaps and dry valleys across the Roum Fault (Fig. 11). The main period of canyon-cutting across the Tyre-Nabatiy6 plateau is likely to be late Messinian in age (c. 5.5 Ma). Consequently, these canyons offer excellent long-term markers of lateral offsets along the R o u m Fault Zone. Most critically, studies in the Hula valley and Galilee (Horowitz 1979) show that the ancestral Litani river flowed axially along the transform. As a consequence, the Litani gorge across the Tyre-Nabatiy6 plateau cannot have been cut by waters flowing from its current catchment. The next drainage basin to the north is that of the modern Zahrani river. If these headwaters lay directly above the lower Litani gorge during late Messinian times, the Roum Fault Zone must have accommodated at least 15 km left-lateral displacement (Fig. l l b ) . The progressive leftlateral offset of the Z a h r a n i headwaters is charted by an array of windgaps and dry gorges between the Z a h r a n i and Litani gorges. However, it seems unlikely to us that the deepest gorge in the Tyre-Nabatiy6 plateau with the most substantial offshore canyon should have developed from such a small catchment

EVOLUTION OF THE LEBANESE RESTRAINING BEND area as is provided by the headwaters of the Zahrani. If the Litani gorge was not cut by the headwaters of the Zahrani, presumably it was by those now lying further north along the eastern side of the Roum Fault (Fig. 11c). A 30 km displacement on the Roum Fault would bring the next catchment area in line with the Litani gorge, that of the Aouali river. The advantages of this restoration (Fig. 11c) are to align the Zahrani headwaters close to windgaps at the head of the Debb6 river. The Damour headwaters align with the Zahrani gorge and the Beirut river aligns with the lower gorge of the Aouali. We conclude that the Roum Fault Zone must have accommodated at least 15 km, probably more than 30 km, left-lateral displacement since the Messinian. Only some of the modern river courses record part of this displacement because the drainage system has been prone to stream capture. This process is evident in the presence of windgaps along the uplifted western flank of the Roum Fault and in the array of dry canyons which cross the Tyre-Nabatiy6 plateau.

Kinematic linkage between the Roum and Dead Sea Fault Systems The structural continuity between faults in southern Lebanon has been obscure (Girdler 1990). The southern part of the Roum Fault is marked by 7-10 km segments which define leftwards stepovers (Fig. 2). This geometry appears to be responsible for adjacent flank uplifts, concentrated on the west side of the fault zone, and associated topographical depressions (Fig. 8). The depressions contain Quaternary coarse clastic deposits, shed from the flanks. Only one sedimentary basin has developed, at Jarmaq (Fig. 8), largely because the low ground is drained by the Litani river and the flank uplift, which otherwise would pond sediment, is breached by the antecedent lower Litani gorge. The left-stepping segmentation on the Roum Fault defines an array of relay ramps (sensu Peacock & Sanderson 1995). The pattern of uplifts and basins associated with this system appears to continue to the northernmost strand of the main Dead Sea Fault System near Metula (Figs 2 and 8). Heimann & Ron (1987) described active faulting on the eastern side of the Hula Valley (Azaz Fault), relaying transtensionally onto the southernmost part of the Hasabaya Fault. This in turn appears to relay across onto the southern part of the Yammouneh. The leftstepping relays generate basins, such as the Quaternary E1 Marj depression (Fig. 8). Rightward bends on the faults, perhaps inherited from

99

when they operated as a linked system with the main Yammouneh Fault, generate minor pressure ridges (e.g. Jebel Hamamiss, Heimann & Ron 1987). Thus the modern transfer of displacement from the Dead Sea Fault System onto the Roum is accomplished by an array of relay ramps, parts of which accommodate transtension and basin formation whereas other parts are associated with transpression and uplift. The segmented nature of the modern Roum Fault and others linking across the the main Dead Sea Fault System in the Hula Valley raises important issues of displacement compatibility and kinematic viability. As active fault segment lengths within the relay ramp zone are between 5 and 10 kin, even displacements on fault strands of 1-2 km would generate very high strains and rotations in the connecting wallrocks. Neither are observed; the Plio-Quaternary basalts record no palaeomagnetic rotations (Ron 1987). This fault array cannot accommodate the much larger displacements estimated for the Roum Fault on the basis of geomorphology or predicted from plate tectonics. We conclude therefore that fault segmentation here is transitory and overprints a through-going, fully linked fault system. It should be noted that transient segment activity must also be a feature of other parts of the Dead Sea Fault System.

The uplift of Mount Lebanon The ages of the major folds are rather less controversial than those of the principal transcurrent faults in Lebanon. The Jabel en Nsour fold belt (Figs 2 and 5) are truncated by the unconformity beneath the Homs Basalt. The Bekaa valley existed as a sedimentary basin during the Miocene, with the accumulation of continental fluvial and lacustrine deposits of this age within the regional syncline. These sediments onlap unconformably the flanks of the Mount Lebanon uplift, particularly to the north of Chtaura (Fig. 2). They also form the fills to valleys emanating from the flanks of the range (Butler & Spencer in press). On the coastal side of the range, transgressive Miocene carbonates, presumably of late Tortonian-Early Messinian age, unconformably overlie discordant Cretaceous strata, notably near Jouni6 (Fig. 2). The ancestral Jabel Barouk was also uplifted at the time, shedding sediment onto the marine terrace now represented by the Tyre-Nabatiy6 plateau (Nammour 1992). That the major uplifts have continued to amplify is indicated by the tilts in Miocene strata away from Mount Lebanon in the Bekaa velley and at Jouni6 (Fig. 2). New folds have grown

100

R.W.H. BUTLER E T A L .

during the Plio-Quaternary. These include the Jabel Turbol anticline and the Kousba fold belt near Tripoli (Fig. 2). These structures have offshore equivalents, for example, the folds at the Ile du Palmier (Beydoun & Habib 1995). Amplification of the coastal monocline is indicated by the array of uplifted marine terraces (Sanlaville 1970; Nammour 1992) throughout coastal Lebanon. That uplift is continuing is indicated by a prominent notch level at about 1 m above sea level between Tyre and Tripoli (Fig. 2). Clearly, folding has occurred throughout the late Miocene-Recent history of the restraining bend. It thus accompanied displacement on the Roum and Yammouneh Faults. They remain the clearest indications of transpression within the Lebanese restraining bend. As discussed above, the major folds are oriented sub-parallel to the Yammouneh and associated faults and therefore show so-called 'in-line' behaviour. Smaller folds, with wavelengths of 1-5 km, can occur in en echelon arrays. Hancock & Atiya (1979) proposed that Lebanese folds generally initiated with E - W trending hinge lines and have subsequently been rotated counter-clockwise during distributed wrench shearing. In contrast, Westaway (1995) suggested that, of all the folds, at least the major anticlines initiated 'in-line', as a result of fault-normal compression. To test between these two models we have analysed the orientation of minor structures with respect to fold hinges on a range of folds. Our rationale was that if folds amplified as foldhinges rotated then minor structures should transect the fold hinge (for models, see Woodcock & Schubert 1994). The range of minor structures within the folds of Lebanon has been described by Hancock & Atiya (1979). Here we present data from part of the major in-line fold of Jabel Barouk (Fig. 12), although the same general patterns have been replicated at every one of the in-line folds so far investigated. The Niha syncline contains a core of upper Cretaceous carbonates between two uplifted tracts of Jurassic carbonates. The larger of the uplifts contains the topographical crest of Jabel Barouk and trends parallel to the Yammouneh Fault. To the north and south of our study area, published maps (e.g. Dubertret 1955) show the Jabel Barouk structure to be an anticline. However, in our study area there is no evidence of an eastern limb to this fold. We will examine this aspect shortly. Minor structures within the Cretaceous carbonates consist of bedding-orthogonal joint sets, bedding sub-parallel stylolites and minor veins. Measurements of these structures, together with bedding measurements to obtain estimates of

v

91

u trot- E o c o ~ o

~ ) ~ J

//

Lwr -M;d C r o t

22

Jurossic

k

L2

Fig. 12. Simplified geological map of the Niha syncline area, part of the in-line fold belt to the Yammouneh Fault in south-central Lebanon (see Fig. 2 for location). The cross-section (X-Y) is a semi-schematic representation of fold geometries adjacent to the fault, showing the Jabel Barouk halfanticline, presumed amplified from the edge of the Mesozoic carbonates on the west side of the fault. The stereonets show poles to bedding (upper one, n = 291) and intersections between bedding and joints or extension veins (lower one, n = 622). fold-hinge orientation, were taken on four transects across the Niha syncline. Poles to bedding define a simple girdle with an average hinge that is essentially horizontal, trending 200 ~. This trend is consistent with the map-pattern (Fig. 12). The lines of intersection between joints or veins and bedding planes show a complex pattern but define a point maximum parallel to the average hinge line to the syncline. These data indicate no sign of transection. We infer therefore that the history of incremental strains (represented by the minor structures) and the finite strain (represented by the major fold) were essential coaxial.

EVOLUTION OF THE LEBANESE RESTRAINING BEND

Relationship between 'in-line' folds and displacement on the Y a m m o u n e h Fault Deformation during transpression is likely to have been partitioned into strike-slip on the major faults, initially the Yammouneh, and fault-normal compression generating folds. The Jabel Barouk structure developed parallel to the Yammouneh Fault. There are sound mechanical reasons for this. Classic studies of buckle folds (e.g. Biot 1961) show that competent layers will preferentially develop fold hinges parallel to their edges, so, if the Yammouneh Fault cut the competent Mesozoic carbonates, the cut itself would have seeded the orientation of the subsequent fold hinge. However, the fold could amplify differentially along the edge of the fault. Half-folds, as shown in our study area, are a common result of such processes (Biot 1961), with fold amplification enhanced because of a locally reduced mechanical resistance to bending. We infer that in this sector the Yammouneh Fault accommodated not only regional left-lateral displacements but also local differential uplift. Certainly, on the eastern flank of Jabel Barouk, there are numerous outcrops of fault breccia that are consistent with our kinematic model. It also provides an explanation for apparently young landscape features along this segment of the fault (Garfunkel et aL 1981) and the current seismicity (Fig. 3) without invoking Plio-Quaternary strike-slip on it. The southern part of the Yammouneh Fault may therefore be behaving as a steep reverse fault whereas the transcurrent displacement is currently accommodated by the Roum Fault Zone.

101

strata (chiefly upper Cretaceous basalts), and their apparently associated strike-slip fault arrays discussed by Ron et aL (1990a). The notion that subsidiary strike-slip faulting and rotation occurred throughout the evolution of the restraining bend may be assessed using the data of Gregor et aL (1974; see Ron et al. 1990a), in conjunction with the structure of their sites. These workers identified counter-clockwise rotations of 55.6~ (+10.4 ~ of Mesozoic volcanic rocks with respect to their expected orientation. Ron et al. (1990a) argued that these rotations require multiple active strike-slip faults to generate these rotations. In flat-lying strata it may be difficult to recognize such faults. However, the sites studied by Gregor et aL (1974) are adjacent to steeply dipping strata. The best example comes from the Laklouk area of Mount Lebanon (Figs 2 and 13). Although the palaeomagnetic sites generally lie

Sonnin~limestone I 3 km I i underlying Cretoceous U~ ~ JUFOSSiC

~"

expected U ~j ~ /direction

~~

Relationship between folds, cross-faults and vertical rotations The conclusion reached above on the basis of minor structures, that the major in-line folds grew coaxially, is in conflict with the notion that the region around the major transcurrent faults such as the Yammouneh experienced a general leftlateral shearing. The general shearing model (e.g. Ron 1987; Ron et aL 1990a) suggests that rotations about vertical axes accompanied by rightlateral slip on subsidiary faults accommodate part of the oblique slip during transpression. In some cases this style of deformation is clear. The 11~ counter-clockwise rotation of Plio-Pleistocene basalts in the Korazim block (described by Heimann & Ron (1993); location shown in Fig. 2) clearly developed synchronously with transcurrent shearing along the major strike-slip faults of the Dead Sea Transform. It is much harder to establish the timing of the larger-scale rotations, based on palaeomagnetic data from Mesozoic

Xt I i -

v=,

.L../

I

t

I

Fill. 13. Simplified geological map and profile through the Qartaba 'horst', showing Gregor et al.'s (1974) palaeomagnetic declinations for Late Jurassic (stippled arrows) and early Late Cretaceous (white arrows) sites. It should be noted that the fold structure of the horst is not disrupted by transverse (E-W) faults. (See Fig. 2 for location.)

102

R.W.H. BUTLER E T A L .

i~"

transform fault zone

}: ~.-~norrow

I! l}I

migrating triple junction ? triple junction r relocate s ~ ~ II

f" /,.:: oblique ~. ,,:~. folds ~ ~ Yommouneh links /: ~ *~;. "" across dlstribufed

transform

"in-line" folds

,

narrow

fault zone

t! A ) !4i''':(~ II~'"'"

/~

ifii

g

,r

distributed trans,o.~

,,

__....,~At',

Yom,ouoeh

..: ;.:,. ,"....".::.:!

tron.form'/

~." ',

/I ~

iocote~ along

9

1\$,1:~~

.one

~

~;

C.

b. 50kin

,,

f

(\~r~,

9~ I

taulting~

through-going ~ f~oumFault / "

/ YA~[ I. . . #./Ji~2s aoanaonea [{l'--transform I .

Jarmoq

I

5kin El Marj basin

Levantine

/-\x d

~. \ I I

,ronstensioo on ~armo, ~ao/,

Nf

~'o I.~ ~"~:~ ~::::!;i!ii :~:basin ;::::i:; ~:::~1

e.

.........

~I

Fig. 14. A model for the kinematic evolution of the Lebanese restraining bend on the Dead Sea Transform, seen in map view. The Palmyrides fold belt (inverted sedimentary basins) is not illustrated but may have been important during the early stages of transform activity. Arabia v. Africa motion is indicated for various stages (white outlined arrows, see also Fig. 1). (a) illustrates a possible early stage of development (c. 18-15 Ma) where the region contains a distributed left-lateral transcurrent shear zone, before nucleation of a throughgoing Yammouneh Fault; (b) shows the main stage of transform activity with the restraining bend represented by the left-lateral Yammouneh Fault and fault sub-parallel folds (Mount Lebanon, Mount Herman-Anti Lebanon); (c) shows the situation with the change in plate convergence vector, causing the restraining bend geometry to geometrically harden and be bypassed by the Roum Fault; (d) shows the modern situation with the Roum Fault becoming segmented with individual fault strands, particularly in the south; (e) shows the resultant geometry of transtension at leftward stepovers but transpression on right-trending portions of faults (fault names in the Hula area after Heimann & Ran (1987)). in gently dipping strata, these sites lie close to a major fold structure, the Qartaba 'horst' (Fig. 13). The eastern limb of the 'horst' is sub-vertical with a N N E - S S W strike. It retains this strike alongside all the Laklouk sites and is not cut by any map-scale faults. Thus, if the palaeomagnetic rotations were accommodated by subsidiary strike-slip faulting, this deformation must entirely precede the development of the Qartaba structure. The rotations therefore formed early in the local structural chronology. As at Laklouk, other sites studied by Gregor et al. (1974) lie adjacent to steeply dipping strata

which define part of the fold structure of Mount L e b a n o n or its southern continuation in the Chouf mountains (Fig. 2). These limbs are only locally cut by cross-faults. The regional extent of fold limbs essentially prohibits any large-scale cross-fault development after folding. Thus if the cross-faults and palaeomagnetically determined rotations are part of the Dead Sea Transform history in the Lebanese restraining bend they must represent a very early stage and have not operated since the Miocene. Gregor et al. (1974) suggested that the rotations affected the whole region and represent large-scale plate

EVOLUTION OF THE LEBANESE RESTRAINING BEND processes entirely preceding the initiation of the transform.

Linking restraining bend evolution to plate tectonics Many existing attempts to link structural evolution within the Lebanese restraining bend to plate kinematics on the Dead Sea Transform have assumed that the array of structures seen in the bend were active contemporaneously. By linking landscape evolution to tectonic structures a picture emerges of evolving deformation. Here we review our findings of structural evolution, schematically illustrated in Fig. 14. In our structural model, the earliest structures within the Lebanese sector of the transform are the oblique folds, cross-faults and tectonic rotations that collectively represent a broad zone of distributed left-lateral shear (Fig. 14a). These field structures form the earliest part of the relative structural chronology. However, the initiation of transform tectonics remains poorly understood, not least because of the difficulties in isolating these effects from those associated with Palmyride deformation (e.g. Chaimov et al. 1990) in the late Paleogene. As noted by Quennell (1984) and others, Palmyride deformation and the D e a d Sea Transform may be kinematically linked, but such an appraisal lies outside the scope of this discussion. However, for Hancock & Atiya (1979), much of the displacement on the Yammouneh post-dates the oblique folds and crossfaults of the Lebanese area. After a period of distributed strain, our model shows the Yammouneh Fault cutting through the Lebanese area to provide a 'hard-linked' through-going transcurrent structure linking the Ghab Fault with the southern part of the Dead Sea Transform (Fig. 14b). This must have occurred in Miocene times, presumably while the Arabia-Africa rotation pole lay far to the west. The position of this pole is critical in determining the amount of transpression across the Yammouneh Fault. The transpression is manifest by coeval transcurrent displacement on the Yammouneh Fault together with increasing uplift and amplification of the Mount Lebanon and related in-line folds. These folds initiated parallel to the main faults and thus record nearideal partitioning of transpressive strain. During the late Miocene the Arabia-Africa rotation pole migrated much closer towards the transform. The effect was to increase the angle of convergence across the Yammouneh Fault. For the restraining bend to remain active, the shortening and amplification rate of in-line folds

103

would have had to increase greatly. Rather, the transform fault zone evolved, with transcurrent displacements transferred off the Yammouneh and onto the Roum Fault system (Fig. 14c). The effect of this reorganization in the plate boundary further north would have been dramatic, with the triple junction between the transform and the Tethyan collision zone migrating from SE Turkey into the NE Mediterranean. There are few studies as yet of structural evolution in this region to confirm our prediction. Nevertheless, the Lebanese sector of the transform remained generally transpressive, as indicated by continued deformation and uplift of Mount Lebanon. Presumably, right-ward steps on the transform existed offshore. In part, the differential uplift of Mounts Lebanon and Hermon relative to the Bekaa was accommodated by steep reverse movements reactivating segments of the abandoned transcurrent faults. This may still be continuing, particularly adjacent to the southern Bekaa area, where there is local seismicity (Van Eck & Hofstetter 1990). For much of the Pliocene the Roum Fault system presumably acted as a through-going transform which accommodated the great bulk of the total 6 mm/a displacement between the Arabian and African plates. It was geometrically well suited to do this, forming a small-circle segment about the Pliocene rotation pole (Quennell 1984; Girdler 1990). However, the focal mechanism studies of Van Eck & Hofstetter (1990; Fig. 1) indicate that the modern rotation pole has moved much closer to the transform. Thus the trend of the Roum Fault system no longer forms a small circle about the active rotation pole and would be strongly in transpression. It appears to have modified, however, into an array of distinct fault segments (Fig. 14d and e). That these segments are generally only a few kilometre long implies that the fault geometries have not accommodated much displacement. We infer that the segments are very young, perhaps only a few thousand years old. In their present configuration (Fig. 14e) the segments step rightwards and are associated with local transtension. This is indicated by an array of relay-ramp basins (Fig. 14e; in the sense of Peacock & Sanderson 1995). However, the western flanks of the fault segments are strongly uplifted and internally folded. Our observations at the Beaufort ridge (Fig. 14e), together with those by Heimann & Ron (1987) along the Shehumit Fault, suggest that these fault flanks are nucleating in-line folds. Thus the individual fault segments are undergoing transpression, presumably because they step leftwards across the active direction of relative plate motion. This

104

R.W.H. BUTLER E T A L .

interpretation is consistent with the modern location of the rotation pole identified in Fig. 1.

Discussion The finite description of the Lebanese restraining bend with its multiple transcurrent faults, cross-faults, tectonic rotations and folds, belies a more complex tectonic history where the different structures record different deformation histories. In this sense, our conclusions amplify those of Heimann & Ron (1993), who described transient fault segmentation. Such variations in fault segmentation may be expected without necessarily invoking any change in plate kinematics. However, the Dead Sea Transform has experienced such variations in long-range kinematics and they are recorded in its Lebanese segment. Although the region displays the finite geometry of a restraining bend, this geometry reflects only part of the tectonic development and transform history. We have been able to build up a view of incremental structural evolution through landscape stratigraphy. Many of the landforms, such as fault scarps, minor offsets of streams and basins at relay ramps, have remained little modified by surface erosional processes long after the cessation of local tectonic activity. Clearly, in the semi-arid, karst-prone environment of Lebanon, many landforms are highly stable (Butler & Spencer in press). Consequently, it is difficult to use satellite images to build a picture of active faults. Similarly, existing geological maps, constructed before modern developments in tectonic geology and without regard to tracing faults, provide incomplete pictures of fault continuity. The following conclusions, based on a reinterpretation of these types of data but enhanced with our own field observations and dating, may require further modification as the structural geology of Lebanon is mapped in its own right. By piecing together local structural evolution within the restraining bend, it is possible to relate the crustal and local lithospheric response to evolving plate kinematics. As a consequence of the migration of the pole of rotation towards the Dead Sea Transform, new fault segments were activated, bypassing the restraining bend and placing part of the crust from transpression into ideal transcurrent or even locally transtensional tectonics. Thus small parts of the fault system, particularly those at the southern end of the restraining bend, have experienced a rather complex geological history over the past 18 Ma, associated with fault array evolution in a rather low-displacement plate boundary system.

Perhaps such complexities should be expected for most continental transforms. A migrating rotation pole represents one of the simplest examples of plate kinematic evolution. The history of, for example, the East Anatolian Transform (Fig. 1) is likely to have been far more complex, with migrating triple junctions and the related, local changes in fault kinematics this process implies. There is a well-established discrepancy between the amount of crustal shortening in the restraining bend compared with its bulk transcurrent displacement (e.g. Hancock & Atiya 1979; Quennell 1984). However, these analyses were made with the Pliocene rotation pole and the total transform displacement of 105 km. We suggest that this discrepancy has been overestimated because the restraining bend geometry (Fig. 14b) was active for only 50-60% of the total displacement on the transform. Furthermore, the Miocene rotation pole implies a much lower angle of relative plate convergence across the restraining bend. Finally, we point out that there are many unsolved problems within the Dead Sea Transform. Although we have presented a structural history for the restraining bend, it is qualitative. Each stage of the history requires quantification of the angular velocities of plate motion about the particular pole and this compared with resolved magnitudes of transcurrent offset, transpression and the observed strains in the field. An incremental approach is required to achieve these goals, with the search for more, datable, landscape features to provide a calibrated tectono-morphological stratigraphy. We thank Z. Beydoun, C. Walley and K. Khair of the American University of Beirut for discussions on Lebanese tectonics, together with M. Casey for discussions on folding. The original draft of this contribution has been improved greatly thanks to the forthright reviews of G. Roberts and two anonymous referees, although the opinions presented here remain the authors'. H. M. G. was supported by a NERC research studentship. R. W. H. B. was supported by a Nuffield Foundation Research Fellowship. K-Ar analyses were performed in Leeds by P. Guise and D. Rex.

References AMBRASEYS,N. N. & BARAZANGI,M. 1989. The 1759 earthquake in the Bekaa valley: implications for earthquake hazard assessment in the eastern Mediterranean region. Journal of Geophysical Research, 94, 4007-4013. ARTHAUD,E, DUBERTRET,L., MASSON,P. ~; MERCIER, J. L. 1978. I~tats de constraintes successifs posteoc6nes les long des failles libano-syriennes au nord de Beyrouth. 6e ROunion Annales Science de

EVOLUTION OF THE LEBANESE RESTRAINING BEND

la Terre, Orsay (en depot d la Societ~ Gdologique de France). BARTOV, Y. 1994. Geological photomap of Israel and adjacent areas at 1:750 000, 2nd edition. Geological Survey of Israel, Jerusalem. BEYDOUN, Z. R. 1977. The Levantine countries: the geology of Syria and Lebanon (maritime regions). In: NAIRN,A. E. M., KAYES,W. H. & STI-mLI,E G. (eds) The Ocean Basins and Margins, 4A, the Eastern Mediterranean. Plenum, New York, 319-353. & HABIB, J. G. 1995. Lebanon revisited: new insights into Triassic hydrocarbon prospects. Journal of Petroleum Geology, 18, 75-90. BLOT,M. A. 1961. Theory of folding of stratified viscoelastic media and its implications in tectonics and orogenesis. Geological Society of America Bulletin, 72, 1595-1620. BUTLER, R. W. H. & SPENCER,S. (in press). Landscape evolution and the preservation of tectonic landforms along the northern Yammouneh Fault, Lebanon. In: WHALLEu W. B., SMITH, B., WIDDOWSON, M. & SUMMERFIELD, M. (eds) Uplift, Erosion and Stability: Geological and Geomorphological Perspectives on Landscape Evolution. Geological Society, London, Special Publication. - - , PRIOR, D. J. & KNIPE, R. J. 1989. Neotectonics of the Nanga Parbat syntaxis, Pakistan, and crustal stacking in the northwest Himalayas. Earth and Planetary Science Letters, 94, 329-343. & GRIFFITHS,H. M. 1997. Transcurrent fault activity on the Dead Sea Transform in Lebanon and its implications for plate tectonics and seismic hazard. Journal of the Geological Society, London, 153, 757-760. CHAIMOV,T. A., BARAZANGI,M., AL-SAAS, D., SAWAF, T. & GEBRAN,A. 1990. Crustal shortening in the Palmyride Fold Belt, Syria, and implications for movement along the Dead Sea Fault System. Tectonics, 9, 1369-1386. CHRISTIE-BLICK,N. & BIDDLE,K. T. (eds) 1985. Strikeslip Deformation, Basin Formation and Sedimentation. Society of Economic Palaeontologists and Mineralogists, Special Publications, 37. CROWELL, J. C. & SYLVESTER,A. G. (eds) 1979. Tectonics of the Junction between the San Andreas Fault System and the Salton Trough, South-eastern California. University of California, Santa Barbara. DRUCKMAN,Y., BUCHBINDER,B., MARTINOTTI, G. M. SIMON TOY, R. & AHARON, P. 1995. The buried Afiq canyon (eastern Mediterranean, Israel): a case study of a Tertiary submarine canyon exposed in Late Messinian times. Marine Geology, 123, 167-185. DUBERTRET, L. 1955. Carte gdologique du Liban au 1/200 000, avec notice explicative. Minist~re des Travaux Publics, Beirut. -& VAUTRIN, M. H. 1945. Carte gOologique du Liban au 1/50 000, feuille Rachaiya-Nord, avec notice explicative. Minist6re des Travaux Publics, Beirut. EYAL, M., EYAL, Y., BARTOV,Y. & STEINITZ,G. 1981. The tectonic development of the western margin -

-

,

-

-

105

of the Gulf of Elat, Aqaba, rift. Tectonophysics, 80, 39-66. FREUND,R. & TARLING,D. H. 1970. Preliminary Mesozoic palaeomagnetic results from Israel and inferences for microplate structure in Lebanon. Tectonophysics, 60, 189-205. ~ - , GARFUNKEL,Z., ZAK, I., GOLDBERG, M., WEISBROD, T. & DERIN, B. 1970. The shear along the Dead Sea rift. Philosophical Transactions of the Royal Society of London, Series A, 267, 107-130. GARFUNKEL, Z. 1981. Internal structure of the Dead Sea leaky transform (rift) in relation to plate kinematics. Tectonophysics, 80, 81-108. & ALMAGOR, G. 1985. Geology and structure of the continental margin off northern Israel and the adjacent part of the Levantine basin. Marine Geology, 62, 105-131. , ZAK, I. & FREUND,R. 1981. Active faulting in the Dead Sea Rift. Tectonophysics, 80, 1-26. GIRDLER, R. W. 1990. The Dead Sea transform fault system. Tectonophysics, 180, 1-13. & SOUTHREN,T. C. 1987. Structure and evolution of the northern Red Sea. Nature, 330, 716-721. GOEDICKE, T. R. 1972. Submarine canyons of the central continental shelf of Lebanon. In: Stanley, D.J. (ed.) The Mediterranean Sea: a Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, PA, 655-670. GREGOR, C. B., MERTZMAN, S., NAIRN, A. E. M. & NEGENDANK, J. 1974. The palaeomagnetism of some Mesozoic and Cenozoic volcanic rocks from the Lebanon. Tectonophysics, 21, 375-395. HALL, J. K., UDINSTEV,G. B. & ODINOKOV,Y. Y. 1994. The bottom relief of the Levantine Sea. In: KRACHENINNIKOV,V. A. & HALL, J. K. (eds) Geological Structure of the North Eastern Mediterranean (Cruise 5 of Research Vessel 'Akademik Nikolaj Strakhov'). Historical Productions Hall, Jerusalem, 5-32. HANCOCK, P. L. & ATIYA, M. S. 1979. Tectonic significance of mesofracture systems associated with the Lebanese segment of the Dead Sea transform fault. Journal of Structural Geology, 1, 143-153. HEIMANN,A. & RON, H. 1987. Young faults in the Hula pull-apart basin, central Dead Sea transform. Tectonophysics, 141, 117-124. & -1993. Geometric changes of plate boundaries along part of the northern Dead Sea Transform: geochronologic and palaeomagnetic evidence. Tectonics, 12, 477--491. --, EYAL, M. & EYAL, Y. 1990. The evolution of Barhta rhomb-shaped graben, Mount Hermon, Dead Sea Transform. Tectonophysics, 180, 101-110. HEMr'rON, M. R. 1987. Constraints on Arabian plate motion and extension history of the Red Sea. Tectonics, 6, 687-705. HIRSCH, F. 1990. Apercu de l'histoire phan6rozoique d'Israel. Journal of" African Earth Sciences, 11, 177-196. HOFSTETTER, A., VAN ECK, Y. & SHAPIRA, A. 1996. Seismic activity along fault branches of the Dead Sea-Jordan Transform System: the Carmel-Tirtza fault system. Tectonophysics, 267, 317-330. -

-

106

R.W.H. BUTLER ETAL.

HOROWITZ, A. 1979. The Quaternary of Israel. Academic Press, New York. JOFFE, S. & GARFUNKEL, Z. 1987. Plate kinematics of the circum Red Sea - - a re-evaluation. Tectonophysics, 141, 5-22. KEMPLER, D. & GARFUNKEL, Z. 1994. Structure and kinematics in the northeastern Mediterranean: a study of an irregular plate boundary. Tectonophysics, 234, 19-32. KHAIR, K., TSOKAS, G. N. & SAWAF,Y. 1997. Crustal structure of the northern Levant region: multiple source Werner deconvolution estimates for Bouguer gravity anomalies. Geophysical Journal International, 128, 605-616. LAWS, E. D. & WILSON, M. 1997. Tectonics and magmatism associated with Mesozoic passive continental margin development in the Middle East. Journal of the Geological Society, London, 154, 459-464. MOR, D. 1987. Geological Map: Har Odem. Geological Survey of Israel. 1993. A time-table for the Levant Volcanic Province, according to K - A t dating in the Golan Heights, Israel. Journal of African Earth Science, 16, 223-234. NAMMOUR, M. T. 1992. Plages a Turrides Mioc6nes dans le Liban-Sud, leurs rapports avec celles du Quaternaire, avec les glacis-cones et les talwegs fossiles, dans le sublittoral libanais. Hannon (R~vue Libanaise de Gdographie), 21, 7-50. NUR, A. & BEN-AVRAHAM, Z. 1978. The eastern Mediterranean and the Levant: tectonics of continental collision. Tectonophysics, 46, 297-311. PEACOCK, D. C. P. & SANDERSON,D. J. 1995. Strike-slip relay ramps. Journal of Structural Geology, 17, 1351-1360. PLASSARD, J. & KOGOL, B. 1981. Seismicitd du Liban. Conseil National de la Recherche Scientifique, Beirut. QUENNELL, A. M. 1958. The structural and geomorphologic evolution of the Dead Sea Rift. Quar-

terly Journal of the Geological Society, London, -

-

114, 1-24. 1984. The western Arabian rift system. In: DKXON,

J. E. & ROBERTSON,A. H. E (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society of London, Special Publications, 17, 775-788. Roy, H. 1987. Deformation along the Yammuneh, the restraining bend of the Dead Sea transform: palaeomagnetic data and kinematic implications. Tectonics, 6, 653-666. -& EYAL,Y. 1985. Intraplate deformation by block rotation and mesostructures along the Dead Sea Transform, northern Israel. Tectonics, 4, 85-105. - - , NUR, A. & EYAL, Y. 1990a. Multiple strike-slip fault sets: a case study from the Dead Sea Transform. Tectonics, 9, 1421-1432. - - & HOFSTETrER, A. 1990b. Late Cenozoic and Recent strike-slip tectonics in Mt. Carmel, Northern Israel. Annales Tectonicae, 4, 70-80. SANLAVILLE,19. 1970. l~tude gOomorphique de la r~gion littorale du Liban. Publications de l'Universit6 Libanaise, Beyrouth. SHAPmA, A. & HOVSTETrER, A. 1993. Source parameters and scaling relationships of earthquakes in Israel. Tectonophysics, 217, 217-226. VAN ECK, T. & HOFSTEaq'ER,A. 1990. Fault geometry and spatial clustering of microearthquakes along the Dead Sea-Jordan rift fault zone. Tectonophysics, 180, 15-27. WALLEY, C. D. 1988. A braided strike-slip model for the northern continuation of the Dead Sea Fault and its implications for Levantine tectonics. Tectonophysics, 145, 63-72. WESTAWAY, R. 1994. Present-day kinematics of the Middle East and eastern Mediterranean. Journal of Geophysical Research, 99, 12071-12090. 1995. Deformation around stepovers in strikeslip faults. Journal of Structural Geology, 17, 831-846. WOODCOCK, N. H. & SCHtmERT, C. 1994. Continental strike-slip tectonics. In: HANCOCK,P. L. (ed.) Continental Tectonics. Pergamon Press, Oxford, 251-263. ZAK, I. & FREUND, R. 1981. Asymmetry and basin migration in the Dead Sea Rift. Tectonophysics, 80, 27-38. -

-

Tectonic evolution of the Northern Salinian Block, California, USA: Paleogene to Recent shortening in a transform fault-bounded continental fragment ENRICO TAVARNELLI

Centro di Geodinamica, Universitd della Basilicata, 85100, Potenzal Italy

Abstract: The complex structural setting of the western margin of North America is interpreted to result from oblique convergence of the North American and Pacific plates, accommodated by both right-lateral slip along the San Andreas fault and shortening east and west of it. Strike-slip movements along the San Andreas fault led to the detachment of a continental fragment (the Salinian Block) from the North American margin during early Miocene, and its translation northwestwards for over 300 km. Structural analysis in the Northern Salinian Block, west of the San Andreas fault, reveals a NE-SW-directed shortening accommodated by NW-SE-trending folds and thrusts with a dip-slip kinematic character. The stratigraphic record of progressively younger unconformities affected by both folds and thrusts, as well as the overprinting relationships among these contractional structures, enables us to clarify the tectonic evolution of the region since Paleocene time. The Paleocene to Recent history of the Salinian Block was dominated by strike-slip along the San Andreas fault, and by shortening perpendicular to it. The partitioning between strikeslip and dip-slip movements appears to be controlled by a pre-existing tectonic feature. The results from structural analysis along the Salinian Block are integrated into a deformation model for the western margin of North America, providing additional constraints on the timing of deformation and helping to clarify the role of strain-partitioning processes in the obliquely convergent California margin. The general parallelism between map- and outcrop-scale structures, known among geologists as Pumpelly's rule (see Price & Cosgrove 1990), renders mesoscopic analysis important for defining the kinematic history of deformed areas. A l t h o u g h outcrop-scale observations have long proven successful in inferring kinematic models (Wilson 1961; Cosgrove 1980), a limit of geodynamic reconstructions solely based on minor structures is that the latter are usually overprinted by later deformations, thus making tectonic inferences problematic. A well-defined geodynamic framework determined by means of independent methods, instead, makes structural analysis a reliable tool for relating detailed field information to larger-scale features. The tectonic setting of the NE circum-Pacific subduction system is characterized by the juxtaposition of the North American plate, largely made of continental lithosphere, to the Pacific, Juan de Fuca and Cocos plates, almost entirely made of oceanic lithosphere. Well-defined seafloor magnetic isochrons, and the presence of two presumably fixed reference frames, the Hawaii and Yellowstone hot-spots, constrain both relative and absolute plate motions (Atwater 1970, 1989; Stock & Molnar 1988; Doglioni & Harabaglia 1996). In central-northern California, the Pacific and North American plates are separated by a continental fragment,

the Salinian Block, where the record of Paleogene to Recent deformations is preserved along superb coastal exposures. These elements, in combination, make central-northern California an unusually favourable setting where mesoscopic data can tentatively be linked to plate-tectonic reconstructions. This paper aims to unravel the deformation history of the Northern Salinian Block, bounded by the active San Andreas fault system to the east, and by the Pacific plate to the west. The results of a structural analysis provide the basis for discussing the role of strike-slip partitioning processes in the tectonic evolution of the North American ocean-continent transform margin.

Geological setting The juxtaposition of the North American and Pacific plates occurs along a belt of anastomosing NW-SE-trending faults, the San Andreas fault system, which extends from the Gulf of California to Cape Mendocino for over 2500 km. Together with its northern continuation, i.e. the Queen Charlotte fault zone, the San Andreas fault system transfers extension produced at the East Pacific Rise to convergence and subduction in the Aleutian Trench with a dominantly rightlateral strike-slip, thus defining a major o c e a n - c o n t i n e n t transform margin (Wilson

TAVARNELLI,E. 1998. Tectonic evolution of the Northern Salian Block, California, USA: Paleogene to recent shortening in a transform fault-bounded continental fragment. In: HOLDSWORTH,R. E., STRACHAN, R. A.& DEWEY,J. E (eds) 1998. Continental Transpressionaland Transtensional Tectonics. Geological Society, London, Special Publications, 135, 107-118.

107

108

E. TAVARNELLI

1965; Atwater 1970). The inception of the transform regime is generally referred to the Late Oligocene (29-26 Ma: Stock & Molnar 1988; Atwater 1989). A diffuse seismicity associated with the San Andreas fault system, outlined by destructive historical and recent earthquakes (Hill et al. 1990), indicates that this regime is still tectonically active. In spite of a general agreement on the role of the San Andreas fault as a major tectonic boundary, controversy exists on the interpretation of its continuation at depth. Wilson (1965), Atwater (1970) and Namson & Davis (1988), among many others, considered it as a deep-rooted fault cutting the entire lithosphere, i.e. the stacked Pacific and North American plates, whereas Jones et al. (1994) and Holbrook et al. (1996) interpreted it as a shallow feature confined to the North American plate and emanating from a 20-25 km deep, gently NE-dipping d6collement. The tectonic history of California before activation of the San Andreas fault system was dominated by the Andean-type convergence with subduction of the Pacific plate under the North American continent, which had been initiated during Late Mesozoic time (Dickinson 1981). Continued subduction gave rise to an arc-trench system, with development of a frontal accretionary prism (the Franciscan Complex), a forearc basin (the seat of deposition of the Great Valley Sequence) floored by oceanic crust (i.e. the future Coast Ranges Ophiolites), and an igneous complex (the Sierra Nevada Magmatic Arc). Subduction-related shortening was probably accompanied by rightlateral slip along an ancestor lineament precursor of the San Andreas fault (the Proto-San Andreas fault: Nilsen 1981; Page 1990), which could have accommodated up to 300 km displacement in pre-Eocene time (Graham & Dickinson 1978). The present-day San Andreas fault in central-northern California separates continental North America from a dominantly granitic fragment, the Salinian Block (Fig. 1). This fragment is thought to represent the former southern continuation of the Sierra Nevada batholith (Graham 1978; Ross 1983), or an exotic terrane accreted to North America during subduction of the Farallon plate (Vedder et al. 1983). Since early Miocene time, the Salinian Block has been detached from the North American margin and has been translated 300 km northwestwards along the San Andreas fault to its present position in western California (Graham & Dickinson 1978; Blake et al. 1984). Four smaller fragments are recognized within the Salinian Block: the Southeastern, Central, Western and Northern

Fig. 1. Sketch map of the Northern Salinian Block, showing location of study areas.

Blocks (Ross 1983). This study will focus on the structural evolution of the Northern Salinian Block, well exposed along the central-northwestern coast of California (Fig. 1). Rocks of the region consist of Lower Cretaceous, mainly granitic and metamorphic basement, unconformably overlain by Upper Cretaceous to Recent sediments. Granitic rocks crop out extensively in the Montara Mountains, in the Point Reyes Peninsula, at Bodega Head, and, together with other igneous and metamorphic assemblages, are believed to underlie the sedimentary sequences of Point Arena and Pigeon Point (Fig. 2). The Upper Cretaceous to Recent mainly marine sequences show abrupt lateral stratigraphic differences over little horizontal distance, which reflect deposition in local, restricted basins, thus suggesting a complex history of continued synsedimentary deformation (Dickinson et al. 1979). Highly simplified

NORTHERN SALINIAN BLOCK, CALIFORNIA

109

Fig. 2. Simplified composite stratigraphic columns of the outcrop locations of the Northern Salinian Block, based on Nilsen et al. (1981), Graham et al. (1989), and references therein. Arrows indicate the deformation episodes documented in this paper.

composite stratigraphic columns of the exposed sediments are summarized in Fig. 2.

General structure Since inception of the transform regime, the deformation history of central-northern California was characterized by dominant rightlateral strike-slip along the San Andreas and attendant San Gregorio, Pilarcitos, Hayward and Calaveras faults (Johnson & Normark 1974; Graham & Dickinson 1978; McLaughlin et al. 1996). Right-lateral movements were accompanied by coeval SW-NE-directed shortening, mainly expressed by NW-SE-trending folds and thrusts, which are abundant east of the San Andreas fault (Aydin & Page 1984; Namson & Davis 1988; Bloch et al. 1993; Jones et al. 1994,

and references therein). By contrast, the internal architecture of the Northern Salinian Block, west of the San Andreas fault, is poorly constrained and structural information is relatively rare, with important exceptions (Galloway 1977; Coppersmith & Griggs 1978; Joyce 1981; Duane Gibson 1983; Wiley & Moore 1983). This section outlines the character, geometry and orientation of mesoscopic (outcrop-scale) and macroscopic (map-scale) contractional structures, i.e. folds and thrusts, which have affected the Northern Salinian Block since the (?)Late Cretaceous-Paleocene interval. Investigated anticlines and synclines range in both wavelength and amplitude from kilometres (i.e. first-order folds) to metres (i.e. third-order folds, according to the definition by Nickelsen (1963)). Deformation occurred under brittle to

110

E. TAVARNELLI

semi-brittle conditions at shallow crustal levels. Two separate tectonostratigraphic terranes are recognized in the Northern Salinian Block (Fig. 3): the La H o n d a Block, between the San Andreas, Pilarcitos and San Gregorio faults, and the Pigeon Point Block, west of the San Gregorio fault.

Pigeon Point Block The structure of the Pigeon Point Block is characterized by both folds and thrusts. Firstorder folds range from open to isoclinal and their wavelengths vary from a few hundred metres to several kilometres. The major mapscale fold is represented by a syncline which affects the Upper Cretaceous Pigeon Point Formation and whose hinge line intersects the coast near Bolsa Point (Fig. 3). This fold trends 305 ~ and plunges 17 ~ towards the NW (Fig. 4a). Second- and third-order folds, parasitic to the Bolsa Point syncline, trend 308 ~ and plunge 19~ towards the NW (Fig. 4a). The Oligocene-Miocene Vaqueros Fm lies unconformably over the Pigeon Point Fm, post-dating the Bolsa

Point syncline, which probably developed during the Paleocene-Eocene interval (see also Joyce 1981). Remnants of the Vaqueros Fm crop out between Franklin Point and Point Afio Nuevo and at Pescadero Beach (Fig. 3), dipping homoclinally 30 ~ towards the N N E (Duane Gibson 1983), and could represent a portion of a WNW-trending map-scale fold. The Vaqueros Fm is unconformably overlain by the Late Pliocene Purisima Fm, which is affected by a first-order anticline-syncline pair whose hinge lines intersect the coast at San Gregorio Beach and Pomponio Beach, respectively. These folds of probable late Pliocene-(?)Pleistocene age, trend NW-SE, subparaUel to the Tertiary Bolsa Point syncline, and plunge gently towards the NW (Fig. 4b). The effects of reverse faulting are particularly evident near Point Afio Nuevo, where mudstones and cherts of the upper Miocene Monterey Fm are thrust southwestwards over Quaternary marine and continental deposits. This fault, known as the Point Afio Nuevo thrust (Coppersmith & Griggs 1978), strikes N W - S E and dips 35~ towards the NE. Mechanical striae and grooves along the main thrust and attendant faults, a weak cleavage in the fault rock and rare synthetic (R) Riedel shears indicate a dip-slip kinematic character. Displacements are towards the SW with a 41 ~ mean slip direction (Fig. 5a). The fault is parallel to bedding in the hanging wall, and makes angles of 20-35 ~ with beds in the footwall, thus defining a hanging wall flat-footwall ramp geometry (Fig. 6a): this pattern indicates a minimum displacement of about 8 m. Other mesoseopic reverse faults, along which the Upper Cretaceous Pigeon Point Fm is thrust over Pleistocene terrace deposits, cropout near Pescadero Beach (Fig. 3). These faults strike N W - S E and dip 20-45 ~ towards both the SW and NE (Fig. 5b). Striae indicate a dip-slip kinematic character and define a 43 ~ mean slip direction (Fig. 5b).

La Honda Block Structural analysis in the northern part of the La Honda Block was carried out in the Montara Mountains, the Point Reyes Peninsula, Bodega Head and Point Arena areas (Fig. 1).

Montara Mountains Fig. 3. Sketch map of the structure of Pigeon Point Block, west of the San Gregorio fault (see inset in Fig. 1); asterisks indicate the outcrop location of described structures.

The main structure recognized in the Montara Mountains is a first-order, asymmetrical NNEverging anticline that is well exposed in the cliffs between Shelter Cove and Devil's Slide (Fig.

NORTHERN SALINIAN BLOCK, CALIFORNIA

~1

b

PigeonPoint

San GregorioPomponio Beach

111

~

9

Devil'sSlides

xL~

[~ . ~

layered gneiss; L>S

~

~ ~

layered gneiss; S = LI ~ ' ~

/

p

o

r/

~ ~ ~ ~ a

~-~ ~

augen gneiss; L>>S ~ ' - - - -

~

~

~

0 d

o

l

a

t

~ -

n

s

o

h

r

t

a

p

h

o

e

s

/

.~--.

t

9-o .\

-

Fsp-augen . \\ " r , v / ~'

"

bundantfolding II L"

e

amphibolite bodies

d

i

3~__~ ~1_ i~,, /:,37/

~~~/-'~'~///

e

/

/

pinch & swell anorthosite sheets

Fig. 6. Schematic block diagram of part of the Nordfjord-Sogn Detachment, Standal. View towards the SE. Not to scale. cleavage is dominant in places. Top-to-the-west extensional crenulation cleavages are abundant within the anorthosite sheets.

Mylonite, chlorite breccia and cataclasite The zone of mylonitic gneiss gradually passes into a zone characterized by both ductile and brittle structures directly below the Standal Fault (NSD) (Figs 5 and 6). This zone consists of dark grey-green to black, fine-grained semipelitic to quartzo-feldspathic schistose greenschist-facies mylonite to ultramylonite with occasional small (millimetre to centimetre) augen of quartz and feldspar and/or a finely spaced lamination defined by S fabrics have also been reported from the VGC (Gilotti & Hull 1993) and S6ranne (1992) reported prolate mylonitic fabrics close below the HCybakken Detachment, just north of the M T F Z (Fig. 1). However, augen gneisses make up only a minor part of the WGR. The bulk of the gneisses in the

East-west trending folds (Fb) within the WGR The folds within the W G R gneisses are generally 1-100 m scale, open to close folds that fold the penetrative amphibolite-facies fabric (Sa) and gneissic layering and overprint the lineation-parallel (Fa) folds described above. Fb fold axes are subparallel to but, in places, fold the regional amphibolite-facies lineation La. Most of the Fb folds are upright but many recumbent folds also occur, in particular on either side of Nordfjord. Commonly recumbent and upright sets occur very close together, without displaying overprinting relationships. The intensity of Fb folding varies greatly; along F0rdefjorden and Nordfjord, the folding led to a north-south shortening of the order of 30-50%. Elsewhere, as on Stadlandet, such folds are open or absent (Fig. 8a). Further north, close to the MTFZ, the fold intensity increases again (S6ranne 1992; Robinson 1995). A detailed down-plunge section has been constructed from excellent and continuous roadcuts between Naustdal and Fcrde (Fig. 8b). Here, two sets of later folds occur, one set dips to the north, the other to the south. No overprinting relationships have been found. Restoration of these folds indicates a n o r t h - s o u t h shortening of 12-29%. Such folding may be expected if contraction was occurring by constriction normal to the fold axes.

EXHUMATION OF UHP ROCKS BY TRANSTENSION

169

N - S SHORTENING N

Slsdllar~iet

0

< 15%

} ,,'- 1~-...-50%,'~,11., NOrdflord ~ , "-e~

~20% ~,~ .....

26~ n

ere

~

H&stelnen

Kvsmhesten

~ ===~~ ~ . ~ . . ~ - - ~

8olund

I soon~f|erd 8 -, ~ I

25km

_

B

-

~

DevonianBasins

~

Western GneissRegion

~7

UFfPeclogites

--

AIIochthonous

a

~ i ~

.

"

N-S shortening: 12-29%

~//,~., L~~176

b

Fig. 8. (a) Section B-B'. North-south cross-section across the WGR, showing approximate amount of north-south shortening during east-west extension. Partly after Kildal (1970) and Osmundsen (1996). North of Hornelen Basin: own observations. (b) Down-plunge projection normal to the stretching lineation (090/30 to E) near Naustdal, showing late folding, coeval with extension.

East-west trending folds in the Middle Allochthon

East-west folding of the Devonian sediments

On the eastern side of the Hornelen Basin, the steep Hyenfjord offers a good opportunity to study east-west trending folding in the Middle Plate between the Hornelen Detachment and the NSD. The Allochthon rocks here consist of gneisses (including augen gneiss), quartzite, anorthosite and minor calc-silicates. Despite its relatively high structural position the Middle Plate has experienced much extension under amphibolite-facies conditions (Wilks & Cuthbert 1994). As a result, many of the extensional structures are very similar to those in the underlying WGR: mylonitic gneiss with top-to-thewest shear sense, a strong east-west lineation, locally intense amphibolite-grade lineationparallel folding and an L >> S fabric in the sporadic augen gneisses. A detailed cross-section along the new road along the eastern shore of Hyenfjord has been made (Fig. 9); outcrops on the steep cliffs on the western shore have been studied with the aid of binoculars from a high viewpoint on the opposite fjord wall. The later folds (Fb) are very well developed along Hyenfjord. Generally, they are close folds with wavelength varying from 10 cm to 10 m and are parasitic to larger open to close folds with a wavelength of 1-4 km. Upright folds are dominant but, locally, 1-10 m scale recumbent folds occur. Total north-south shortening is estimated to be a minimum of 18%. Along Hyenfjord, the Fb folds are cut by the Hornelen Detachment; this can be clearly observed in the field below the GrCndalen Syncline (Fig. 9).

East-west folding also affects the Devonian basins, especially the Hornelen and Kvamshesten Basins. The Kvamshesten Basin and the NSD are folded in a single upright, kilometrescale syncline about the same size as the basin itself. North-south shortening across the Kvamshesten Basin is estimated to be 11-25% (Osmundsen 1996). In the Hornelen Basin, several kilometre-scale, open to close folds occur with most folding occurring close to the north and south margins, for instance the Grcndalen syncline in the south (Fig. 9). Bedding is locally vertical or overturned, as along the northern margin. The plunge of the folds is 20-30 ~ towards the east together with the overall east-dipping Devonian strata. The total north-south shortening of the Hornelen Basin along the Hyenfjord section is estimated to be about 12%. The GrCndalen Syncline and the folds near Skjerdal are cut by the Hornelen Detachment, indicating that north-south shortening occurred before the latest stages of top-tothe-west displacement (Fig. 9).

East-west folding of the Detachments The NSD is folded in east-west folds with wavelength of about 10 km, the Devonian basins generally occurring in the synclines (Figs 2 and 8). The total north-south shortening of the NSD as measured from Hornelen to Solund is c. 20-25 % (Fig. 8). The Hornelen Detachment, however, is an almost flat structure, with less than 2% north-south shortening along Hyenfjord (Fig. 9).

170

M. KRABBENDAM & J. F. DEWEY

S

Hyen

N

Nordfjord

Gronclilen Syncline .,, ~ : ~

o~ o o : f=o oo ~ Hornelen ~ ~176 o 0~.~ ~176176 ~o~ o0 ~ o~=_;.~0~ o0 o0 o~ ~ o o 00 ~

Basin

o-,., -

C

Middle AIIochthonous mylonites

Linear a u g e n gneiss

2%

C w

with top to w e s t shear sense

Fig. 9. Section C-C'. South-north cross-section of eastern end of the Hornelen Basin, showing different amounts of north-south shortening of the Hornelen Basin, Hornelen Detachment (HD) and Middle-Lower Allochthons. Structural data in the Allochthons based on detailed fieldwork; structural data from Devonian and the HD based on binocular observation and reconnaissance fieldwork. Southern part of section partly after Bryhni & Lutro (1989). The northern part of the section as drawn is extrapolated from outcrops farther west, where the northern contact of the Hornelen Basin is exposed. (See Fig. 2 for location.)

This suggests that the Hornelen Detachment is a younger structure than the NSD and the mylonites in both the W G R and the Middle Plate.

Relative age o f north-south shortening The relative age of the north-south shortening resulting from the east-west trending folds (Fb) has been somewhat controversial and has been connected with the elusive Solundian-Svalbardian Orogeny (e.g. Torsvik et al. 1988). The relationship between north-south shortening and Devonian deposition is, as yet, also unclear. Bryhni & Skjerlie (1975) argued for basin formation during north-south shortening for the Kvamshesten Basin. This was further supported by Chauvet & S6ranne (1994), who claimed that all the east-west folds of the Devonian rocks and the detachment were syn-depositional, based on different types of unconformities as interpreted from aerial photographs of the Kvamshesten and Hornelen Basins (S6ranne et al. 1989). Fieldwork by Osmundsen (1996) has not confirmed the existence of the unconformities in the Kvamshesten Basin. Osmundsen (1996) argued that most of the north-south contraction in the Kvamhesten Basin is late- to post-depositional. On the other hand, reconnaissance fieldwork near the southern margin of the Hornelen Basin in Grcndalen strongly suggests that the complementary anticline south of the Grcndalen Syncline contains an angular unconformity on its north limb and is syn-depositional. The Hyenfjord cross-section (Fig. 9), however, displays the relevant field relationships,

confirming the mapping of Bryhni & Lutro (1989). The Hornelen Detachment cuts the Grcndalen Syncline and other folds in the Devonian rocks near Skjerdal. Regardless of whether these folds are syn-depositional (which is not of particular relevance here), they clearly pre-date substantial movement along the Hornelen Detachment. The south bounding fault of the Hornelen Basin cuts the Hornelen Detachment (Bryhni & Lutro 1989) and is, locally, steeper than bedding and, therefore, post-dates both deposition and extensional movement along the Hornelen Detachment. Significantly, the north-south shortening by Fb folds of the gneissic layering (>18%) exceeds the Fb north-south shortening of the Hornelen Basin (c. 12%), which in turn exceeds the Fb north-south shortening of the Hornelen Detachment (c. 2%) (Fig. 9). This suggests the following evolution for the Hyenfjord area: (1) folding of gneissic layering and mylonites, including the underlying NSD; (2) deposition of Devonian basin, at that moment positioned farther east; (3) folding of Devonian basin; (4) emplacement of Devonian basin along the Hornelen Detachment; (5) further folding of the Hornelen Detachment, the Hornelen Basin and the underlying gneisses (about 2%). This evolution strongly suggests folding, possibly progressive, before and during Devonian sedimentation and during top-to-the-west movement along the Hornelen Detachment. Progressive folding, before and after faulting along the NSD, was also suggested by Torsvik et al. (1986) for the Kvamshesten Basin. We suggest that north-south shortening affecting the Devonian basins is related in time to the regional east-west

EXHUMATION OF UHP ROCKS BY TRANSTENSION

171

extension, indicating bulk constriction. In summary, north-south contraction was synchronous with vertical thinning and east-west extension for a considerable period during the late-orogenic evolution of the WGR, indicating late-orogenic bulk constriction.

excision or attenuation from the Hornelen Basin to the U H P province on Stadtlandet is at least 80 km, this implies that at least 80% of the exhumation was achieved by bulk constriction and only 20% by excision along the NSD (see also Krabbendam & Wain 1997).

Orientation of lineations and fold axes

Interpretation and discussion

Lineations and fold axes, derived from various sources, show a regional swing in the plunge direction within the W G R (Fig. 1). To the south, the lineations are W N W - E S E plunging, in the central part the plunge direction is east-west, whereas close to the MTFZ the lineations and fold axes are NE-SW, subparallel to the MTFZ. This swing in orientation affects both Fa and Fb structures. Fb orientations also swing in areas where close or tight Fa structures are absent, so that the influence of pre-existing corrugations of Fa on the orientation of Fb can be regarded as insignificant.

To exhume the U H P and HP rocks in the W G R in Norway, some 100 km of overburden must have been removed from a restricted area (the U H P province), and some 50-60 km of overburden over a large area (>10 000 km2). The widespread occurrence of extensional structures, the very significant metamorphic break across the NSD, the scarcity of Devonian sediments in the foreland regions of the Scandinavian Caledonides and the presence of intramontane Devonian basins above extensional detachments indicate that the W G R was exhumed mainly by extensional tectonics with only a minor contribution by erosion. To explain the constrictional fabrics, the syn-extensional north-south shortening and the Devonian strike-slip movement along the MTFZ we present below a model of sinistral transtensional exhumation in the WGR.

Strain estimation In the absence of stratigraphic or other displacement markers, strain is hard to estimate in the WGR. In the western part of the WGR, vertical shortening has been estimated by Dewey et al. (1993) to be about 0.2 (80%) on the basis of flattening of foliation around eclogite pods and shortening of granite and quartz veins in Sunnfjord, and by assuming (from eclogite metamorphic assemblages) an original crustal thickness of c. 150 km that has been reduced to the present c. 30 km. The linear feldspar augen at Standal have aspect ratios of 1 : 1 - 2 : 5 - 1 0 , indicating minimum vertical shortening of 0.1-0.2 (80-90%), concomitant with similar horizontal north-south shortening. As described, the north-south shortening responsible for the Fb structures varies between y -- z, K = oo) occurs at a transtensional angle of 70.5~ at all other transtensional angles 'general constriction' occurs (x > 1 > y > z, 1 < K < ~). At angles smaller than 70.5 ~ vertical shortening exceeds horizontal shortening (z is vertical) whereas, at angles greater than 70.5 ~, horizontal shortening exceeds vertical shortening (y is vertical) (Fig. 10). In sinistral transtension, the transport direction (TD) is anticlockwise to the finite stretching direction (x), which is in turn

172

M. KRABBENDAM & J. F. DEWEY ~.

transport

(180-t8)>70.5~

8

Fig. 10. Volume-constant, homogeneous, boundary-compatible transtension. Transtensional angle is (180 -13). (a) Transtensional angle >70.5~ horizontal shortening exceeds vertical shortening and z is horizontal. (b) Transtensional angle S fabrics form parallel to x (finite stretching direction). Pinch-and-swell structures may develop at low transtensional angles parallel to the xy-plane with long axes parallel to x. Both recumbent and upright folding are expected with axes parallel to x. Late dykes and normal faulting may occur normal to xi. Block rotation associated with faulting will have axes of rotation normal to x. A crustal cross-section normal to the finite stretching direction x (Fig. 12) shows that during bulk transtension the crustal thickness can be reduced dramatically, coeval with bulk horizontal shortening normal to x. If a low-angle extensional detachment zone develops (vorticity partitioning in the xz plane) it must fold and extend as it moves (Fig. 12). There is a strong tendency for the detachment to lock itself, not only as it increases its surface area per unit volume as a result of folding but also because of the vertical rotation of the finite stretching direction with progressing deformation, so that subsequent displacement has to

move over, rather than along, the hinges. Rejuvenation of detachments is, therefore, to be expected (Fig. 12).

Partially partitioned, sinistral transtension in the W G R To explain the structures described in this paper, the following exhumation model for the W G R is proposed (Figs 13 and 14). Transtension in the W G R was sinistral with the strike-slip component parallel to the MTFZ (NE-SW). The transtension was partially partitioned, with an increase in transtensional angle towards the NW. The increasing transtensional angle is indicated by the regional swing in the mineral lineations and fold axes' azimuths from N W - S E close to the Jotun Detachment, W N W - E S E around Solund, east-west in the study area to W S W - E N E to S W - N E close to the MTFZ (Fig. 1). Variable transtensional angles are indicated also by the variable intensity of the late east-west folds as described above (Fig. 8).

174

M. KRABBENDAM & J. F. DEWEY

Fig. 13. Schematic model of partially partitioned sinistral transtension in West Norway. (See text for explanation.) JD, Jotun Detachment; HD, HCybakken Detachment; LGFZ, L~erdal-Gjende Fault Zone; MTFZ, MCre-TrCndelag Fault Zone; NSD, Nordfjord-Sogn Detachment; R, R0ragen Basin; RD, RCragen Detachment; VGC, Vestranden Gneiss Complex; WGR, Western Gneiss Region.

EXHUMATION OF UHP ROCKS BY TRANSTENSION The transtensional angle was close to zero (near plane strain extension) near the present east edge of the Jotun Nappe (Fig. 13), as indicated by the approximate down-dip, top-to-theNW movement along the Jotun Detachment and the Lerdal-Gjende Fault Zone (Milnes et al. 1988, 1997; Fossen & Holst 1995) and by limited lineation-parallel folds in this area, suggesting limited horizontal shortening normal to the extension direction. In the study area around the Devonian basins, the transtensional angle is c. 40-50 ~, compatible with the values of the strain estimation of bulk-constriction. Towards the MTFZ, the transtensional angle approaches 90~ (near strike-slip) towards the MTFZ (Fig. 11) supported by intense, upright folding and linear fabrics, subparallel to the MTFZ, in this area (see below). In an early, hot stage and/or deeper part of the orogen, transtension was taken up by relatively homogeneous constriction, producing linear fabrics in augen gneisses and the finely spaced gneissic layering and the tight, lineation-parallel folding (Fa) of this layering. In a later, cooler, shallower part of the orogen, strong vorticity partitioning in the xz-plane took place, with the development of the distinct extensional NSD in the upper crust. The NSD folded while it was moving to the west, and rejuvenation of the detachment took place to the north of the H~steinen Basin, thus developing the Hornelen Detachment. In this late stage, north-south partitioning of the transtension was possibly more pronounced with the development of distinct NE-SW trending strike-slip faults, such as the sinistral MTFZ and subsidiary sinistral strikeslip fault zones. It was in the cooler parts of the orogen that conspicuous mylonite zones (both underlying the NSD and elsewhere in the WGR) were developed, although the bulk of the extension was taken up by the more homogeneously distributed strain of the earlier, hotter phase. The syn-tectonic low-grade metamorphism of the Devonian basins above the W G R can be explained by the juxtaposition of these basins on top of warm lower crust along the NSD and the Hornelen Detachment; the cleavage formation within the basins was caused by the north-south shortening accompanying transtension. Within the Hitra and Smr Devonian basins, positioned along the MTFZ, syn-metamorphic (sub-greenschist-facies) structures include upright and recumbent ENE-WSW trending folds, with associated cleavages (Bee et al. 1989). Locally, two cleavages are developed, with the younger cleavage anticlockwise to the older cleavage (Bee et al. 1989). These structures are compatible with partitioned transtension near the MTFZ, although the structural evolution in

175

this area may be more complex because of the suggested docking and rotation of the Smr terrane (Torsvik et al. 1989). Transtension north o f the M T F Z : the Vestranden Gneiss C o m p l e x

North of the MTFZ, a similar partitioned transtension took place with the transtensional angle increasing south-eastward towards the MTFZ (Fig. 13). Close to the MTFZ, the Vestranden Gneiss Complex (VGC, Fig. 1) is characterized by upright NE-SW trending folds, (sub)-parallel to the lineation with rare new axial surface fabrics (Gilotti & Hull 1993). The gneisses show NE-SW trending, amphibolitefacies L-S and L > S fabrics. Numerous smallscale steep shear zones show ductile sinistral strike-slip movement (Gilotti & Hull 1993). Constrictional fabrics also occur below the Hcybakken Detachment (Fig.l, S6ranne 1992). All these structures are very similar to those observed in the western part of the WGR. In the NW of the VGC, the Roan Window (Fig. 1) contains HP granulite (T -- 870~ P = 14 kbar (Johannson & M611er 1986; MOiler 1988), with peak metamorphism dated at 432 _+ 6 Ma (Dallmeyer et al. 1992), i.e. early Scandian. The Einardsdalen D6collement Zone (EDZ) separates the high-grade Roan Window in the footwall from a lower-grade hanging wall with deformation structures developed at granulite facies but continuous fabric formation at lower T and P (Gilotti & Hull 1993). Although interpreted as a thrust by Gilotti & Hull (1993), the above features are more easily explained by extension rather than thrusting (Lister & Davis 1989; Wheeler & Butler 1994). The shear sense of the Einardsdalen D6collement is to the NW (Gilotti & Hull 1993), which is more compatible with the regional extension direction than the direction of overthrusting, which is to the SE (Fossen & Rykkelid 1992). We suggest, therefore, that the E D Z is a top-to-the-NW extensional detachment and not a thrust, and that the structures observed by Gilotti & Hull (1993) can be better explained by partitioned, sinistral transtension, and reject their transpressional interpretation. In the VGC area, close to the MTFZ, the transtensional angle was high, resulting in strongly constrictional strain (Fig. 13). Towards the NW, the transtensional angle decreases and is low in the area around the Roan Window, dominated by top-to-the-NW, near-plane strain extension. Thus the partitioning from low to high transtensional angle towards the MTFZ was more or less symmetrical, occurring on either side of the MTFZ.

176

M. KRABBENDAM & J. E DEWEY

increasing strike slip component (increasing transtensionalangle) ~o~e ~t

1

2

~ -

~~-I

_

I x - - - ~ t "-, _~_L~'.~.,..~\ "

~

orthogonal stretching component

~

Fig. 14. Schematicdiagramof partially partitioned transtension(volumeconstant,boundarycompatible). Partitioning of transtensJona]angleonly; orthogonalstretchingis kept constant. Transtension and oblique plate divergence between Laurentia and Baltica Of interest is whether the transtension in the WGR is associated with oblique, sinistral plate divergence between Laurentia and Baltica. Fossen (1992) argued that late-orogenic extension was accompanied by plate divergence, based on the very widespread evidence of lateorogenic extension in SW Norway and on the systematic overprint of contractional structures by extensional structures along the Jotun Detachment. The syn- or ?late-orogenic Ringerike Group sediments in the Oslo area are thrust towards the east (Bjcrlykke 1983); this thrusting, however, is not balanced by the amount of extension in the WGR and Central Norway, again suggesting plate divergence (Fossen 1992, 1993). Oblique plate divergence is supported also by the reinterpretation of the Caledonian sole thrust in central East Greenland as a top-to-theeast extensional detachment (Hartz & Andresen 1995), which separates Precambrian gneisses, reworked during the Caledonian, from Late Proterozoic and Carnbro-Ordovician meta-sediments. Above the detachment a small but deep Devonian basin occurs, bounded to the west by the Western Fault Zone, which is interpreted by some workers (Larsen & Bengaard 1991) as sinistral during the Devonian. It appears, therefore, that many of the late-orogenic, Devonian deformational features in the East Greenland Caledonides can be explained by sinistral transtension, which was strongly partitioned between extensional detachments and strikeslip faults. This suggests that the late-orogenic evolution of the Caledonides was relatively symmetrical. This study shows that the plate divergence

between Baltica and Laurentia during the Devonian must have been sinistrally oblique as the north-south shortening structures and the east-west constrictional fabrics in the WGR are not compatible with either orthogonal plate divergence or with convergent orogenic collapse. We suggest that Devonian plate divergence in Scandinavia was highly oblique and that transtension was the most important factor in the exhumation of the UHP rocks in Norway. Silurian convergence and collision of Laurentia and Baltica was also sinistrally oblique (Soper et al. 1992), as suggested by sinistral transpression in the Greenland Caledonides (Holdsworth & Strachan 1991), SE Scandian thrusting in the Scandinavian Caledonides (Fossen 1992) and large-scale sinistral movements between Laurentia and Avalonia in the British Caledonides (Dewey & Shackleton 1984; Hutton 1987; Hutton & McErlean 1991). We propose that this sinistral oblique convergence in the Silurian was followed by sinistral oblique divergence in the Devonian. Thus, the change in plate motion vectors between Baltica and Laurentia was less than 180~ possibly not exceeding 45~. The structures in the WGR, presented in this paper, were formed mainly under amphibolitefacies conditions, relatively late in the orogenic evolution. There is a gap in the structural record of about 5-10 Ma, between the formation of eclogite-facies structures (Andersen et al. 1991, 1994; Dewey et al. 1993) and the formation of the late-orogenic amphibolite-facies structures described in this paper (Fig. 3b). During this time it is possible that some form of orogenic collapse may have taken place, quickly followed by the transtension, for which the evidence is presented here. Whether such a collapse was caused by removal of the thermal boundary

EXHUMATION OF UHP ROCKS BY TRANSTENSION layer (Dewey 1988; Platt & England 1993) or by slab break-off (von Blanckenburg & Davies 1995) is uncertain, although some kind of such process is very likely to have taken place. Although the best argued examples of orogenic collapse happen to occur in plate convergent regimes (the Betic Cordillera (Platt & Vissers 1989) and the Himalayas (Dewey 1988; England & Houseman 1988), there is no reason to assume that orogenic collapse and convective removal of the lithospheric root cannot be accompanied by (or even result in) plate divergence.

Oblique plate divergence as an effective exhumation m e c h a n i s m This study shows that oblique plate divergence, expressed by bulk transtension over a wide area, can be a very effective mechanism to exhume a large (U)HP terrane. Various criteria can be put forward to recognize this mechanism in other collision belts: (1) Late-orogenic extension must exceed late-orogenic shortening. (2) Lateorogenic extension is likely to affect the entire crustal thickness. (3) Constriction, represented either by bulk-constrictional fabrics or by folded detachments, and other strain features characteristic of transtension are expected if plate divergence is oblique. (4) Age and structural position of HP units. HP m e t a m o r p h i s m is expected to be peak orogenic and exhumation is expected to be late orogenic and not followed by more convergence. As a result, HP units are likely to be the lowest tectonostratigraphic unit, with HP metamorphism relatively late in the orogenic evolution. If such criteria are not met, HP units are likely to be exhumed by other mechanisms, such as underplating and extension, erosion, extrusion or orogenic collapse during plate convergence. Subduction roll-back, which may lead to an extensional regime where many of the above criteria are met, evidently needs a subducting oceanic crust relatively close to the exhumed HP terrane. The observation of Platt (1993) that exhumation usually occurs while convergence is active, may be biased, as this observation is mainly based on studies of active or young mountain belts. We feel that although the statement may be true for many belts, its generalization is not appropriate. In other words, in 100-200 Ma, the Himalayas and the Alps may well look very much like the WGR, with the largest HP unit (currently still residing at depth) being exhumed by plate divergence, with some tiny, early orogenic HP units occurring at higher structural levels, exhumed while convergence was still active. In this context, it is worth noting

177

that most of the large exposed (U)HP terranes in the world are no longer in convergence.

Conclusions Transtension was the d o m i n a n t e x h u m a t i o n mechanism for (U)HP rocks in the WGR. This study shows that transtension is a powerful exhumation mechanism. Transtension in the W G R was partially partitioned with increasing transtensional angle towards the M T F Z and requires active oblique plate divergence between Laurentia and Baltica. A change in plate motion between Laurentia and Baltica from sinistral oblique convergence to sinistral oblique divergence took place at the end of the Silurian-Early Devonian. It remains uncertain whether this change in plate motion was associated with or even triggered by TBL removal or slab break-off. T. B. Andersen is thanked for many discussions and for taking M. K. up to critical exposures around the Hornelen Basin by helicopter. A. Wain is thanked for many discussions on the metamorphic evolution of the WGR. B. Sturt and an anonymous reviewer are thanked for comments on the manuscript. NERC funding for M. K. is acknowledged.

References ANDERSEN, T. B. & JAMTVEIT,B. 1990. Uplift of deep

crust during orogenic extensional collapse: a model based on field studies in the Sogn-Sunnfjord region of Western Norway. Tectonics, 9, 1097-1111. , JAMTVEIT,B., DEWEY,J. F. & SWENSSON,E. 1991. Subduction and eduction of continental crust: major mechanisms during continent-continent collision and orogenic extensional collapse, a model based on the south Norwegian Caledonides. Terra Nova, 3, 303-310. --, OSMUNDSEN, P. T. & JOLIVET, L. 1994. Deep crustal fabrics and a model for extensional collapse of the southwest Norwegian Caledonides. Journal of Structural Geology, 16, 1191-1203. --, SKJERLIE, K. P. & FURNES, H. 1990. The Sunnfjord Melange, evidence of Silurian ophiolite accretion in the West Norwegian Caledonides. Journal of the Geological Society, London, 147, 5%68. BERRY, H. N., Lux, D. R., ANDRESEN,A. & ANDERSEN, T. B. 1994. Argon 40-39 dating of rapidly uplifted high pressure rocks during late-orogenic extension in southwestern Norway. Geological Society of America, Abstracts' with Programs, 25 6, A-477. BJf~RLYKKE, K. 1983. Subsidence and tectonics in late Precambrian and Palaeozoic sedimentary basins in Southern Norway. Norges Geologiske UndersOkelse Bulletin, 380, 159-172.

178

M. K R A B B E N D A M & J. F. DEWEY

BOUNDY, T. M., ESSENE, E. J., HALL, C. M., AusTRHEIM, H. & HALLIDAY, A. N. 1996. Rapid exhumation of lower crust during continent-continent collision and late extension: evidence from 40Ar/39Ar incremental heating of hornblendes and muscovites, Caledonian orogen, western Norway. Geological Society of America Bulletin, 108, 1425-1437. - - , MEZGER, K. & ESSENE, E. J. 1997. Temporal and tectonic evolution of the granulite-eclogite association from the Bergen Arcs, western Norway. Lithos, 39, 159-178. BOE, R., ATAKAN,K. & STURT,B. A. 1989. The style of deformation in the Devonian rocks on Hitra and Sm01a, Central Norway. Norges Geologiske Undersokelse Bulletin, 414, 1-19. BREKKE, H. & SOLBERG, P. O. 1987. The geology of Atl0y, Sunnfjord, Western Norway. Norges Geologiske UndersOkelse Bulletin, 410, 73-94. BRYHNI, I. 1966. Reconnaissance studies of gneisses, ultrabasites, eclogites and anorthosites in Outer Nordfjord, Western Norway. Norges Geologiske UndersOkelse Bulletin, 241, 1-68. & ANDRI2ASSON, P. G. 1985. Metamorphism in the Scandinavian Caledonides. In: GEE, D. G. & STURT, B. A. (eds) The Caledonide OrogenScandinavia and Related Areas. John Wiley, New York, 763-782. & SKJERLIE, F. J. 1975. Syndepositonal tectonism in the Kvamshesten district (Old Red Sandstone), western Norway. Geological Magazine, 112, 593-600. -& LUTRO, O. 1989. Naustdal. Berggrunnsgeologisk kart 1218 III, scale 1 : 50 000. Forelcpig utgave Norges geologiske unders0kelse. & STURT, B. A. 1985. Caledonides of southwestern Norway. In: GEE, D. G. & STURT, B.A. (eds) The Caledonide Orogen - Scandinavia and Related Areas. John Wiley, New York, 89-107. CHAUVET, A. & DALLMEYER, R. D. 1992. 4~ mineral dates related to Devonian extension in the southwestern Scandinavian Caledonides. Tectonophysics, 210, 155-177. & St~RANNE,M. 1989. Microtectonic evidence of Devonian extensional westward shearing in SW Norway. In: GAYER, R. A. (ed.) The Caledonide Geology of Scandinavia, Graham & Trotman, London, 245-254. & -1994. Extension-parallel folding in the Scandinavian Caledonides: implications for lateorogenic processes. Tectonophysics, 238, 31-54. , KIENAST, J. R., PINARDON, J. L. & BRUNEL, M. 1992. Petrological constraints and PT path of Devonian collapse tectonics within the Scandian mountain belt (Western Gneiss Region, Norway). Journal of the Geological Society, London, 149, 383-400. COTKIN, S. J., VALLEY, J. W. & ESSENE, E. J. 1988. Petrology of a margarite-bearing metaanorthosite from Seljeneset, Nordfjord, western Norway: implications for the P-T history of the Western Gneiss Region during Caledonian uplift. Lithos, 21, 117-128. CUTHBERT, S. J. 1991. Evolution of the Devonian -

-

-

-

-

-

-

-

-

-

Hornelen Basin, Western Norway: new constraints from petrological studies of metamorphic clasts. In: MORTON, A. C., TODD, S. P. & HAUGHTON, P. D. W. (eds) Developments in, Sedimentary Provenance Studies. Geological Society, London, Special Publications, 57, 343-360. DALLMEYER, R. D. 1990.4~ mineral age record of a polyorogenic evolution within the Seve and Kr nappes, Tr0ndelag, Norway. Tectonophysics, 179, 199-226. & GEE, D. G. 1986.4~ mineral dates from retrogressed eclogites within the Baltoscandian miogeocline: implications for a polyphase Caledonian orogenic evolution. Geological Society of America Bulletin, 97, 26-34. , JOHANSSON, L. & MOLLER, C. 1992. Chronology of Caledonian HP-granulite-facies metamorphism, uplift and deformation within northern parts of the Western Gneiss Region. Geological Society of America Bulletin, 104, 444-455. DEWEY, J. F. 1988. Extensional collapse of orogens. Tectonics, 7, 1123-1139. & SHACKLETON,R .M. 1984. A model for the evolution of the Grampian tract in the early Caledonides and Appalachians. Nature, 312, 115-121. , RYAN, P. D. & ANDERSEN, T. B. 1993. Orogenic uplift and collapse, crustal thickness, fabrics and metamorphic phase changes: the role of eclogites. In: ALABASTER, H .M., HARRIS, N. B. W. & NEARY, C. R. (eds), Magmatic Processes and Plate Tectonics, Geological Society, London, Special Publications, 76, 325-343. DIETLER, T. N., KOESTLER, A. G. & MILNES, A. G. 1985. A preliminary structural profile through the Western Gneiss Complex, Sognefjord, southwestern Norway. Norsk Geologisk Tidsskrift, 65, 233-235. D OBRZHINETSKAYA,L. F., EIDE, E. A., LARSEN,R. B., ET AL. 1995. Microdiamond in high-grade metamorphic rocks of the Western Gneiss region, Norway. Geology, 23, 597-600. DRANSFIELD, M. W. 1994. Extensional exhumation of high-grade metamorphic rocks in Western Norway and the Zanskar Himalayas. PhD thesis, University of Oxford. DUNNING, G. R. & PEDERSEN, R. g. 1988. U/Pb dating of ophiolites and arc-related plutons of the Norwegian Caledonides: implications for the development of Iapetus. Contributions to Mineralogy and Petrology, 98,13-23. ENGLAND, P. C. & HOUSEMAN, G. 1988. The mechanics of the Tibetan Plateau. Philosophical Transactions of the Royal Society of London Series A, 326, 301-320. FOSSEN, H. 1992. The role of extensional tectonics in the Caledonides of south Norway. Journal of Structural Geology, 14, 1033-1046. 1993. The role of extensional tectonics in the Caledonides of south Norway: Reply. Journal of Structural Geology, 15, 1381-1383. -& HOLST, T. B. 1995. Northwest-verging folds and the northwestward movement of the Caledonian Jotun Nappe, Norway. Journal of Structural Geology, 17, 3-15. -

-

-

-

E X H U M A T I O N OF U H P ROCKS BY TRANSTENSION & RYKKELID, E. 1992. Postcollisional extension of the Caledonide orogen in Scandinavia: structural expressions and tectonic significance. Geology, 20, 737-740. FURNES, n . PEDERSEN, P. B., SUNDVOLL, B., TYSSELAND, M. & TYMYR, O. 1989. The age, petrography, geochemistry and tectonic setting of the late Caledonian GhsCy Intrusion, west Norway. Norsk Geologisk Tidsskrifi, 69, 273-289. GEBAUER, D., LAPPIN, M. A., GRUNENFELDER, M. & WYTTENBACI-I, A. 1985. The age and origin of some Norwegian eclogites: a U - P b zircon and R E E study. Chemical Geology, 52, 227-247. GEE, D. G. 1978. Nappe displacement in the Scandinavian Caledonides. Tectonophysics, 47, 393-419. , LOBKOWICZ, M. & SINGH, S. 1994. Late Caledonian extension in the Scandinavian Caledonides - the RCragen Detachment revisited. Tectonophysics, 231, 139-155. GILOTTI, J. A. & HULL, J. M. 1993. Kinematic stratification in the hinterland of the central Scandinavian Caledonides. Journal of Structural Geology, 15, 629-646. GRIFFIN, W. L. & BRUECKNER,H. K. 1980. Caledonian Sm-Nd ages and a crustal origin for Norwegian eclogites. Nature, 285, 319-321. & -1985. REE, Rb-Sr and Sm-Nd studies of Norwegian eclogites. Chemical Geology, 52, 249-271. , AUSTRHEIM, H. BRASTAD, K., ET AL. 1985. High-pressure metamorphism in the Scandinavian Caledonides. In: GEE, D. G. & STURT, B. A. (eds) The Caledonide Orogen - Scandinavia and Related Areas. John Wiley, New York, 783-802. GRONLIE, A. & ROBERTS, D. 1989. Resurgent strikeslip duplex development along the Hitra-Smr and Verran Faults, MCre-Trondelag Fault Zone, Central Norway. Journal of Structural Geology, 11, 295-305. , NAESNER, C. W., NAESER, N. D., MITCHELL, J. G., STURT, B. A. & INESON, P. R. 1994. Fissiontrack and K - A r dating of tectonic activity in a transect across the MCre-TrCndelag Fault Zone, central Norway. Norsk Geologisk Tidsskrift, 74, 24-34. HARTZ, E. & ANDRESEN, A. 1995. Caledonian sole thrust of central East Greenland: a crustal-scale Devonian extensional detachment? Geology, 23, 637-640. & ANDERSEN, T. B. 1994. Structural observations adjacent to a large scale extensional detachment zone in the hinterland of the Norwegian Caledonides. Tectonophysics, 231, 123-137. HEINRICH, C. A. 1982. Kyanite-eclogite to amphibolite facies evolution of hydrous mafic and pelitic rocks, Adula Nappe, Central Alps. Contributions to Mineralogy and Petrology, 81, 30-38. HOLDSWORTH, R. E. & STRACHAN,R. A. 1991. Interlinked system of ductile strike-slip and thrusting formed by Caledonian sinistral transpression in NE Greenland. Geology, 19, 510-513. HOSSACK, J. R. 1984. The geometry of listric growth faults in the Devonain basins of Sunnfjord, W. -

-

-

-

,

-

-

179

Norway. Journal of the Geological Society, London, 141, 629-637. HurroN, D. H. W. 1987. Strike-slip terranes and a model for the evolution of the British and Irish Caledonides. Geological Magazine, 124, 405-425. & MCERLEAN, M. 1991. Silurian and Early Devonian sinistral deformation of the Ratagain Granite, Scotland; constraints on the age of Caledonian movements on the Great Glen fault system. Journal of the Geological Society,

London, 148, 1-4. JAMTVEIT,B. 1987. Metamorphic evolution of the Eiksundal eclogite complex, Western Norway, and some tectonic implications. Contributions to Mineralogy and Petrology, 95, 82-99. , CARSWELL, D. A. & MEARNS, E. W. 1987. Chronology of the high-pressure metamorphism of Norwegian garnet peridotites/pyroxenites. Journal of Metamorphic Geology, 9, 125-139. JOHANSSON, L. d~;MOLLER, C. 1986. Formation of sapphirine during retrogression of a basic high-pressure granulite, Roan, Western Gneiss Region, Norway. Contributions to Mineralogy and Petrology, 94, 29-41. KILDAL, E. S. 1970. Geologisk kart over Norge, berggrunnskart. Mhloy, 1:250 000. Norges Geologiske UndersCkelse. KOENEMANN, F. H. 1993. Tectonics of the Scandian Orogeny and the Western Gneiss Region in southern Norway. Geologische Rundschau, 82, 696-717. KRABBENDAM, M. &; WAIN, A. 1997. Late-orogenic structures, differential retrogression and structural position of HP and U H P rocks in the Nordfjord Stadtlandet area, Western Gneiss Region. Norges Geologiske UndersOkelse Bulletin, 432, 127-139. KROC;H, E. J. & CARSWELL,D. A. 1995. HP and U H P eclogites and garnet peridotites in the Scandinavian Caledonides. In: COLEMAN,R. C. & WANG, I. (eds) Ultra High Pressure Metamorphism. Cambridge University Press. KULLERUD, L., TORUDBAKKEN, B. O. & ILEBEKK, S. 1986. A compilation of radiometric age determinations from the Western Gneiss Region, South Norway. Norges Geologiske UndersOkelse Bulletin, 406, 17-42. LAPPIN, M. A. 1966. The field relationships of basic and ultrabasic masses in the basal gneiss complex of Stadtlandet and Almklovdalen, Nordfjord, southwestern Norway. Norsk Geologisk Tidsskrift, 46, 439-496. LARSEN, P. H. & BENGAARD, H. J. 1991. Devonian basin initiation in East Greenland; a result of sinistral wrench faulting and Caledonian extensional collapse. Journal of the Geological Society, London, 148, 355-368. LISTER, G. S. & DAVIS, G. A. 1989. The origin of metamorphic core complexes and detachment faults formed during Tertiary continental extension in the northern Colorado River region, U.S.A. Journal of Structural Geology, 11, 65-94. Lux, D. R. 1985. K/Ar ages from the Basal Gneiss Region, Stadlandet area, Western Norway. Norsk Geologisk Tidsskrift, 65, 227-286.

180

M. K R A B B E N D A M & J. F. DEWEY

MILNES, A. G., DIETLER, T. N. & KOESTLER, A. G. 1988. The Sognefjord northshore log - a 25 km depth section through Caledonized basement in western Norway. In: KRISTOFFERSEN, Y. (ed.) Progress in Studies of the lithosphere in Norway. Norges Geologiske UndersCkelse Special Publication, 3, 114-121. - - , WENNBERG, O. P., SKAR, • & KOESTLER,A. G. 1997. Contraction, extension and timing in the South Norwegian Caledonides: the Sognefjord transect. In: BURG, J. P & FORD, M. (eds) Orogeny through Time. Geological Society, London, Special Publications, 121, 123-148. MOLLER, C. 1988. Geology and metamorphic evolution of the Roan area, Vestranden, Western Gneiss Region, Central Norwegian Caledonides. Norges Geologiske UndersOkelse Bulletin, 413, 1-32. MORK, M. B. E. 1985. A gabbro to eclogite transition on FlemsCy, SunnmCre, western Norway. Chemical Geology, 50, 283-310. & MEARNS, E. W. 1986. Sm-Nd isotopic systematics of a gabbro-eclogite transition. Lithos, 19, 255-267. - - . , KULLERUD, K. & STABEL, A. 1988. Sm-Nd dating of Seve eclogites, Norrbotten, Sweden evidence for early Caledonian (505 Ma) subduction. Contributions to Mineralogy and Petrology, 99, 344-351. NORTON, M. G. 1987. The Nordfjord-Sogn Detachment, W. Norway. Norsk Geologisk Tidsskrift, 67, 93-106. OSMUNDSEN, P. T. 1996. Late-orogenic structural geology and Devonian basin formation in Western Norway: a study from the hanging wall of the Nordfjord-Sogn Detachment in the Sunnf]ord Region. Doctor Scientiarum thesis, University of Oslo. & ANDERSEN, T. B. 1994. Caledonian compressional and late-orogenic extensional deformation in the Staveneset area, Sunnfjord, Western Norway. Journal of Structural Geology, 16, 1385-1401. PLATr, J. P. 1986. Dynamics of orogenic wedges and the uplift of high-pressure metamorphic rocks. Geological Society of America Bulletin, 97, 1037-1053. 1993. Exhumation of high-pressure rocks: a review of concepts and processes. Terra Nova, 5, 119-133. -& ENGLAND, P. C. 1993. Convective removal of lithosphere beneath mountain belts: thermal and mechanical consequences. American Journal of Science, 293, 307-336. & VISSERS,R. L. M. 1989. Extensional collapse of thickened continental lithosphere: a working hypothesis for the Alboran Sea and Gibraltar Arc. Geology, 17, 540-543. ROBERTS, D. 1983. Devonian tectonic deformation in the Norwegian Caledonides and its regional perspectives. Norges Geologiske UndersOkelse Bulletin, 380, 85-96. ROBINSON, P. 1995. Extension of Trollheimen tectonostratigraphic sequence in deep synclines near

-

-

-

-

-

-

Molde and Brattvfig, Western Gneiss Region, southern Norway. Norsk Geologisk Tidsskrifl, 75, 181-198. ROYDEN, L. & BURCnFIEL, B. C. 1987. Thin-skinned extension within the convergent Himalayan region: gravitational collapse of a Miocene topographic front. In: COWARD, M. P., DEWEY, J. F. & HANCOCK, P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 611-619. SEGURET, M., S~;RANNE,M., CHAUVET,A. & BRUNEL, M. 1989. Collapse basin: a new type of extensional sedimentary basin from the Devonian of Norway. Geology, 17, 127-130. SERANNE, M. 1992. Late Palaeozoic kinematics of the MCre-Trr Fault Zone and adjacent areas, central Norway. Norsk Geologisk Tidsskrift, 72, 141-158. -& SEGURET, M. 1987. The Devonian basins of western Norway: tectonics and kinematics of an extending crust. In: COWARD, M. P., DEWEY, J. F. & HANCOCK, P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 537-548. - - , CHAUVET,A. ,St;GURET, M. & BRUNEL,M. 1989. Tectonics of the Devonian collapse-basins of Western Norway. Bulletin de la Soci~tO GOologique de France, 8, 489-499. SMITH, D. C. 1984. Coesite in clinopyroxene in the Caledonides and its implication for geodynamics. Nature, 310, 641-644. 1995. Microcoesites and microdiamonds in Norway: a brief review. In: COLEMAN, R. C. & WANG, I. (eds) Ultra High Pressure Metamorphism. Cambridge University Press. & LAPPIN, M. A. 1989. Coesite in the Straumen kyanite-eclogite pod, Norway. Terra Nova, 1, 47-56. SOPER, N. J., STRACHAN, R. A., HOLDSWORTH, R. E., GAYER, R. A. & GREILING, R. O. 1992. Sinistral transpression and the Silurian closure of Iapetus. Journal of the Geological Society, London, 149, 871-880. STEEL, R. J. & GLOPPEN, T. G. 1980. Late Caledonian (Devonian) basin formation, western Norway: signs of strike-slip tectonics during infilling. In: BALLANCE, P. E & READING, H. G. (eds) Sedimentation in Oblique-slip Mobile Zones. Special Publication of the International Association of Sedimentologists, 4, 79-104. - - , SIEDLECKA,A. & ROBERTS,D. 1985. The Old Red Sandstone basins of Norway and their deformation: a review. In: GEE, D. G. & STURr, B. A. (eds) The Caledonide Orogen - Scandinavia and Related Areas. John Wiley, New York, 293-316. TORSVIK, T. H., STURT,B. A., RAMSAY,D. M., BERING, D. & FLUGE, P. R. 1988. Paleomagnetism, magnetic fabrics and the structural style of the Hornelen Old Red Sandstone, Western Norway. Journal of the Geological Society, London, 145, 413-430. - - , ET AL. 1989. Palaeomagnetic constraints on the early histroy of the MCre-TrCndelag Fault Zone, Central Norway. In: KISSEL,C. & -

-

,

,

E X H U M A T I O N OF U H P ROCKS BY TRANSTENSION LAJ, C. (eds) Paleomagnetic Rotations and Continental Deformation. Kluwer, Dordrecht, 431-457. , , , KISCH, H. J. & BERING, D. 1986. The tectonic implications of Solundian (Upper Devonian) magnetization of the Devonian rocks of Kvamshesten, western Norway. Earth and Planetary Science Letters, 80, 337-347. , SWENSSON,E., ANDERSEN,T. B. & DEWEY, J. F. 1992. Paleomagnetic dating of fault rocks: evidence for Permian and Mesozoic movements and brittle deformation along the extensional Dalsfjord Fault, western Norway. Geophysical Journal International, 109, 565-580. VAN ROERMOND, H. L. M. 1985. Eclogites of the Seve Nappe, central Scandinavian Caledonides. In: GEE, D. G. & STURT,B. A. (eds) The Caledonide Orogen - Scandinavia and Related Areas. John Wiley, New York, 293-316. VON BLANCKENBURG, F. & DAVIES, J. H. 1995. Slab

181

breakoff: a model for syncollisional magmatism and tectonics in the Alps. Tectonics, 14, 120-131. WAIN, A. 1997. New coesite-eclogite occurrences in western Norway: the nature of an ultrahigh pressure province in the Western Gneiss Region. Geology, 25, 927-930. WHEELER, J. 1991. Structural evolution of a subducted continental sliver: the northern Dora Maira massif, Italian Alps. Journal of the Geological Society, London, 148, 1101-1113. & BUTLER, R. W. H. 1994. Criteria for identifying structures related to true crustal extension in orogens. Journal of Structural Geology, 16, 1023-1027. WILKS,W. J. & CUTHBERT,S. J. 1994. The evolution of the Hornelen Basin detachment system, western Norway: implications for the style of late orogenic extension in the southern Scandinavian Caledonides. Tectonophysics, 238, 1-30.

The Trans Mojave-Sierran shear zone and its role in Early Miocene collapse of southwestern North America R O Y K. D O K K A 1, T I M O T H Y

M. R O S S 2 & G A N G L U 3

1Department of Geology and Geophysics, Louisiana State University, Baton Rouge, L A 70803, USA 2Department of Geological Sciences, California State University, San Bernardino, CA 92407, USA 3Marathon Oil Company, P.O. Box 3128, Houston, T X 77253-3128, USA Abstract: The relative motion between the Pacific and North American plates in early Miocene time was not parallel to the overall NW strike of the transform, but was instead oblique and transtensional. It has recently been proposed that in response to this divergence, the western edge of the North American plate east of the transform gravitationally collapsed and moved 100-150 km to the southwest ($50~176 The region of collapse covered an area of nearly 106 km 2 and included what is now southern California, southwestern Arizona, and northwestern Mexico. A major structure facilitating collapse between 21 and 18 Ma was the Trans Mojave-Sierran shear zone (TMSSZ). This east-west shear zone linked the classic detachment fault terranes and metamorphic core complexes of the Mojave desert, southeastern California, southern Arizona, and Sonora, Mexico, to the transtensional plate boundary. To more fully understand the nature and kinematic significance of the TMSSZ and its role in facilitating early Miocene fragmentation of the North American plate, a palinspastic reconstruction of the Mojave desert was performed to remove the disruptive effects of the TMSSZ and younger tectonic events. Features formed just before movements along the TMSSZ were used as markers to assess the TMSSZ deformation. Our analysis indicates that TMSSZ deformation was distributed across a c. 90 km wide band; restoration of markers to their original positions implies that >80 km of dextral shear occurred along the TMSSZ. First-order dextral shear deformation within the TMSSZ is expressed internally by clockwise vertical axis rotations of large areas that were facilitated by second-order zones of sinistral shear that separated the blocks. These second-order sinistral zones apparently exploited older transfer zones of the 24-21 Ma Mojave Extensional Belt.

Plate tectonic principles have been widely and successfully applied to explain Neogene deformation of coastal regions of the s o u t h e r n Cordillera (e.g. Atwater 1970; Ingersoll 1982; Stock & Molnar 1988; Severinghaus & Atwater 1990; Nicholson et al. 1994; B o h a n n o n & Parsons 1995; Fig. 1). Early attempts to apply these principles to explain the distribution and origin of the classic extensional terranes within the interior of the N o r t h A m e r i c a n plate were thwarted because of unresolved uncertainties of older plate models and by the lack of clear-cut geometric and kinematic links to the global plate circuit. Recent plate tectonic reconstructions (Stock & Molnar 1988; Atwater 1989; Severinghaus & Atwater 1990) have improved our comprehension of the uncertainties and behaviour of the P a c i f i c - N o r t h A m e r i c a n portion of the global plate circuit, and have resulted in bringing us closer than ever to fulfilling the promise of fuller tectonic u n d e r s t a n d i n g set forth in Atwater's (1970) plate tectonic tour de force.

The nagging mechanical question of direct, physical linkage of the global plate circuit with c o n t e m p o r a n e o u s early Miocene tectonic systems of the southern Cordillera (e.g. extension in southern California-Arizona, the San Andreas fault system, the Sierran orocline) has been apparently solved by the recognition of the Trans Mojave-Sierran shear zone (TMSSZ; Fig. 1), a broad (c. 90 km), approximately E - W dextral shear zone that passes t h r o u g h the s o u t h e r n Sierra N e v a d a region and Mojave desert (Dokka & Ross 1995; this paper). This link was detected and identified from regional analysis of field structural and palaeomagnetism data. D o k k a & Ross proposed that the TMSSZ formed along the northern edge of a large fragment of the North American plate that detached in early Miocene time in response to transtension developed along the Pacific-North American plate boundary. In addition to providing the kinematic linkage between the global plate circuit and c o n t e m p o r a n e o u s early Miocene

DOKKA, R. K., ROSS,T. M. & LU, G. 1998. The Trans Mojave-Sierran shear zone and its role in Early 183 Miocene collapse of southwestern North America. In: HOLDSWORTH,R. E., STRACHAN,R. A. & DF~WEY,J. E (eds) 1998. Continental Transpressionaland Transtensional Tectonics. Geological Society, London, Special Publications, 135, 183-202.

184

R.K. DOKKA E T A L .

Fig. 1. Index map of the present-day positions of tectonic features and localities of the southwestern USA and northern Mexico that are mentioned in the text. Plate tectonic elements from Severinghaus & Atwater (1990). Dark grey areas, early Miocene detachment terranes and metamorphic core complexes of southeastern California and Arizona. Blackened areas, early Miocene Mojave Extensional Belt (including offset portion in the Salinian Block of central California (Dokka 1989)). Ruled area, Trans Mojave-Sierran shear zone (TMSSZ); sense of shear shown by arrows. Lightly shaded area, late Neogene Eastern California shear zone (ECSZ). Fp, Farallon plate; Mp, Monterey plate; S, Salinian Block; SCT, Santa Catalina and Tortolita Mountains; SM, South Mountain; WM, Whipple Mountains; F.Z., fracture zone; T.J., triple junction; state abbreviations standard. tectonic systems of the southern Cordillera (e.g. extension in southern California-Arizona, the San Andreas fault system, the Sierran orocline), the TMSSZ has been important in rearranging the position of older palaeogeographical and tectonic elements. This paper presents the results of a structural analysis of the Mojave desert that was performed to gain insights into the original geometry and kinematic history of the TMSSZ, and its role in the early Miocene transtensional

collapse of southwestern North America. Our study consisted of two steps. Before we could analyse the structure of the TMSSZ, we first needed to restore the Mojave desert back to its early Miocene (c. 18 Ma) configuration, just after the end of movement along the TMSSZ. This required that subsequent translations, rotations, and strains associated with the 0-13 Ma Eastern California shear zone be accounted for and restored (Fig. 2). U p o n reaching this point, we used well-constrained 24-21 Ma markers to

THE TRANS MOJAVE-SIERRAN SHEAR ZONE

185

Fig. 2. The Pacific-North American transform boundary in the western USA highlighting the location of the Eastern California shear zone (lightly shaded; modified from Dokka (1993)). observe and measure the deformational effects of the TMSSZ.

Early Miocene palinspastic resonstruction of the Trans Mojave-Sierran shear zone The need f o r reconstruction It is fundamental to the concept of structural analysis that the effects of younger events in an area be sequentially removed before one can accurately assess the nature of older events. In intricately deformed areas such as the Mojave desert, this requires that all translations, vertical-axis rotations, and strains associated with each event be known and explained. As will be discussed below, several models for Mesozoic and Cenozoic tectonics of the Mojave desert region are incorrect because they fail to account for the disruptive effects of all Miocene and younger deformations. During late Cenozoic time, the Mojave experienced three structurally different and temporally separated intervals of deformation: (1) approximately N-S directed opening of the c. 24-21 Ma Mojave Extensional Belt (Dokka 1989; Ross 1995); (2) approximately E - W striking, dextral shearing (c. 21-18 Ma) along the TMSSZ (Dokka & Ross 1995, 1996; this paper); (3) the c. 13-0 Ma Eastern California shear zone (Dokka & Travis 1990a, b; Dokka 1993). Subsequent to its time of activity at 21-18 Ma, the TMSSZ was truncated and disrupted by the 0-13 Ma Eastern California shear zone (ECSZ (Fig.

2); D o k k a & Travis 1990a; D o k k a 1993). Approximately 65 km of slip (resolved along an approximately N40~ line) has occurred within this c. 80 km wide belt of distributed righ t shear since its inception (Dokka & Travis 1990a); Pezzopane & Weldon (1993) proposed that the ECSZ continues through northern California and Nevada to eastern Oregon, where it may connect with the Cascade Range. Faults of the southern portion of the ECSZ join with those of the San Andreas system near the Pinto Mountain fault and final merger is completed in western Sonora, Mexico. The physical connection of faults of the ECSZ with the San Andreas fault system demonstrates that the ECSZ is a key element of the Pacific-North American plate boundary; the ECSZ has accommodated, and continues to accommodate, between 9% and 23% of the total relative plate motion (Sauber et al. 1986, 1994; Dokka & Travis 1990b; Savage et al. 1990). Given these disruptive effects, palinspastic reconstruction of the region is thus essential before meaningful structural analysis of the TMSSZ can be carried out.

Methodology Reconstruction of the Mojave desert region to remove the effects of the 0-13 Ma ECSZ followed the protocol established by Dokka & Travis (1990a) and Dokka (1993). The improved reconstruction presented here, as well as the original Dokka & Travis (1990a) model, is constructed primarily to explain regional, two-dimensional, surface (x,y) relations; vertical strain implications of the ECSZ such as local crustal extension and

186

R.K.

DOKKA

ETAL.

.., /

..~L~)

/.

D,! i t

~":

">"

/

,'

!7

.

...

/ .-"

I

.;

;~.

I/~ I I

.,: '~'" , ,

' ~.,, 9 "

e,,,.'" " ~

~-~'Z :-,.

I~.-.

I ""' >' ~ ..:: " .' ..i~ , ( 7 / ..'.~ "., " ' ""~. ". ~" ' , ~:/~".,~-I :'.. ."4~ ,., " ~ " ' ~, ', s. ." , . - . x"s "r . . . .

i

. ::

,':"

~ , , ; ~ +-

.', ~