structural geology of robben island : implications …crowe/downloads/rowe_etal_2010.pdf · south...

C.D. ROWE, N.R. BACKEBERG, T. VAN RENSBURG, S.A. MACLENNAN, C. FABER,

C. CURTIS AND P.A. VIGLIETTI57

IntroductionThe Saldania Belt is one of several deformed belts of“Pan-African” age recording the closure of theAdamastor Ocean with the construction of southwesternGondwana (Frimmel and Fölling, 2004), collectivelyreferred to as the Brasiliano Orogeny (e.g. Cawood andBuchan, 2007, and references therein) or AdamastorOrogeny (Goscombe and Gray, 2008). In southernAfrica, these belts are represented by the Kaoko Belt inNamibia, the Gariep Belt in Namibia/South Africa, andthe Saldania Belt in South Africa. Dextral transpressionalong northwest-striking, east-ramping thrusts has beensuggested for parts of the Gariep Belt (Hälbich and Alchin, 1995), where sheath folds also suggest a strong component of shear-parallel extrusion. The Kaoko Belt displays highly oblique sinistral

shortening (Goscombe et al., 2003; Goscombe and Gray,2008). The Saldania Belt has not been subject to thesame level of structural investigation, due largely torelatively poor exposure.

As noted by Rozendaal et al. (1999), the deformationof the Saldania Belt is not consistent with a ”propercollisional orogeny”. The Saldania Belt is represented inthe Western Cape by two terranes: the inland, slightlyhigher grade Swartland Terrane and the coastalMalmesbury Terrane (Figure 1; Belcher and Kisters,2003). Belcher and Kisters (2003) gave a detailedoverview of structural fabrics in the Swartland Terrane(northeast of the Colenso Fault, Figure 1) and showedthat transposed fabrics therein were formed by highshear strains. Belcher (2003) suggested that a subductiontrench active during the complex, multi-phase closure of

STRUCTURAL GEOLOGY OF ROBBEN ISLAND :IMPLICATIONS FOR THE TECTONIC ENVIRONMENT OF SALDANIANDEFORMATION

C.D. ROWEDepartment of Geological Sciences, University of Cape Town, Private Bag X3, Rondebosch 7701,South AfricaNow at: Department of Earth and Planetary Sciences, University of California, Santa Cruz, 1156 HighStreet, Santa Cruz, CA 95064, United States of Americaemail: [email protected]

N.R. BACKEBERG, T. VAN RENSBURG, S.A. MACLENNAN, C. FABER, C. CURTISAND P.A. VIGLIETTIDepartment of Geological Sciences, University of Cape Town, Private Bag X3, Rondebosch 7701,South Africaemail: [email protected]; [email protected]; [email protected];[email protected]; [email protected]; [email protected]

© 2010 March Geological Society of South Africa

ABSTRACT

We present a detailed structural and lithologic map of Robben Island, offshore Cape Town, South Africa. Robben Island is underlain

by the Tygerberg Formation, part of the Neoproterozoic to early Cambrian Malmesbury Group of the Saldania Belt.

The depositional setting and structural history of the Tygerberg Formation are poorly constrained due to limited outcrop and lack

of previous structural studies.

Sedimentary structures are indicative of deposition at relatively high rates in a high energy environment and we concur with

previous workers that deposition occurred on turbidite fan systems in a tectonically deepening basin. By comparison with active

and ancient examples, we suggest that a forearc or trench slope, supra-subduction zone basin is a possible match to the setting of

the Tygerberg Formation. However, limits on preservation and insufficient age data prevent comparisons in basin geometry and

deposition rates which could be used to test depositional setting with more certainty.

Northwest-southeast striking subvertical pressure solution cleavage is pervasive throughout the exposures. Upright folds, with

axial planes parallel to the cleavage, plunge 10 to 15° to the northwest or southeast with approximately 20° variation in trend

azimuth. The folds are limited in along-axis extent and often occur in asymmetric pairs. Subtle bedding-parallel shear zones divide

folds of different plunge directions. This pattern of folds is consistent with experiments and observations of en echelon folding

during distributed strain associated with oblique transpression. This finding is consistent with previous studies of parallel, slightly

earlier orogenic belts to the north (Gariep and Kaoko Belts) although our observations do not allow us to distinguish whether

transpressional strain was sinistral or dextral. Sinistral transpression is considered more likely given the dominantly sinistral strike-

slip history on the nearby Colenso Fault and the southward migration of collision along the western margin of Africa during the

late Neoproterozoic to early Cambrian.

SOUTH AFRICAN JOURNAL OF GEOLOGY, 2010, VOLUME 113.1 PAGE 57-72

doi:10.2113/gssajg.113.1-57

SOUTH AFRICAN JOURNAL OF GEOLOGY

STRUCTURAL GEOLOGY OF ROBBEN ISLAND58

Figure 1. Regional geologic map, Cape Town area, South Africa. Redrawn from Belcher and Kisters (2003) with additional geology from

Scheepers (1995). Sources for igneous ages: 1 = Scheepers and Armstrong (2002); 2 = Scheepers and Poujol (2002); 3 = Da Silva et al. (2000);

4 = Jordaan et al. (1995); 5 = Armstrong et al. (1998). Da Silva et al. (2000) also reported an age of 536 ± 5 Ma for the Robertson I-type

granite (east of the map area) which is the only recent age data available for the I-type suite. Additional previous dates by Rb-Sr or Pb-Pb

methods are not used in this study. Inset: study location (grey box) relative to African continent. Borders of South Africa, Namibia, Swaziland

and Lesotho are shown.


C. CURTIS AND P.A. VIGLIETTI


59

the Adamastor Ocean could provide an appropriatemodel in which deposition of deep-marine sediments,and structures relating to high-shear strain, could beroughly co-eval and therefore a good match to hisobservations of the Swartland Terrane. The ColensoFault is a poorly exposed strike-slip fault with a complexstrain history comprising a reversal from sinistral todextral motion during rapid regional exhumation at 540 Ma (Kisters et al., 2002). The Malmesbury Terrane,mostly represented by the Tygerberg Formation, hasnever been the subject of a similarly detailed andregionally extensive structural study.

With this contribution we aim to document thedetailed structures within the Tygerberg Formationexposed on Robben Island and compare thesedimentary facies and structures to the proposedenvironments of deposition and deformation.

Geologic setting of Robben IslandRobben Island is a World Heritage site of tremendousimportance to South Africa in terms of both human andnatural history. The coastline of the island (Figure 1)provides a unique opportunity to observe the structuralfabrics of the Neoproterozoic to Cambrian Saldania Belt. The tectonic history of the Saldania Belt is poorly understood. The belt is composed of low-grade (low greenschist and below) metasediments of the Malmesbury Group (Hartnady et al., 1974). The depositional age is not well constrained but a fewdetrital zircon dates suggest that the sediments rangefrom 750 to 560 Ma (Armstrong et al., 1998). Rozendaalet al. (1999) described the Saldania Belt rocks as marine sediments deposited in a deepening basin. The Malmesbury Group was divided by Hartnady et al.(1974) into three tectonic terranes separated by theColenso Fault and the (inferred) Piketberg- WellingtonFault. Detailed field mapping by Belcher and Kisters(2003) revised the stratigraphy by demonstrating that theeastern terranes (Swartland and Boland Terranes of Hartnady et al. (1974)) are overlain, either structurallyor unconformably, by the western Malmesbury Terrane,represented in coastal outcrops by the TygerbergFormation. The Swartland Terrane is deformed by broadregional doubly-plunging anticlinoria, with highermetamorphic grade (up to biotite-bearing schists) in thecenter of the folds. Most of the Swartland Terrane ischlorite-white mica schists and psammites with strongshear fabrics transposing bedding, overprinted byhorizontal shortening. The Malmesbury Terrane isdominated by fabrics suggesting significant horizontalshortening. Although the contact is nowhere exposed,Belcher and Kisters (2003) and Belcher (2003) inferred adepositional unconformity based on the strong shearfabrics of the Swartland Group which are absent fromthe Malmesbury Group, the presence of Swartland-derived vein quartz clasts in the Piketberg formationwhich forms the base of the Malmesbury Group, and thehorizontal shortening which represents the latestdeformation of both groups.

The Saldania Belt is intruded by the Neoproterozoicto Cambrian Cape Granite Suite (Figure 1). Complicatingstructural interpretations, these units were later involvedin the Carboniferous-Permian Cape Orogeny (Frimmel et al., 2001). The Cape Granites are spatially separatedinto a seaward belt of locally tectonised S-type granitesand a landward belt of undeformed I-type granites(Scheepers, 1995). Available dates are suggestive of asouthward-younging trend in the S-type granitesalthough the resolution of the ages is not sufficient toconfirm this trend (Figure 1). Post-tectonic A-typegranites and ignimbrites occur in both terranes(Scheepers, 1995). The granites and their intrusivecontacts appear to be undeformed during the CapeOrogeny, and they clearly crosscut pre-existing foldsand foliations in the Saldania Belt meta-sediments (Von Veh, 1982). Therefore, it has been generallyassumed that the majority of structural development inthe Saldania Belt can be attributed to pre-intrusiondeformation. This assumption is difficult to test however,as the fold axes and regional shortening fabrics areparallel in the Saldania Belt metasediments and theoverlying Paleozoic Cape Basin, and the maximummetamorphic grade at the present level of exposure inboth cases is lower greenschist or lower. There is little clear evidence of structural or metamorphicoverprinting during the second orogeny, which may beused to interpret the “pre-Cape” condition of theSaldania Belt.

Von Veh (1982) studied detailed structural andsedimentological transects across the intrusive contactbetween the Cape Granite and the Tygerberg Formationat Sea Point, Cape Town. He established that theTygerberg sediments were folded prior to graniteintrusion at Sea Point, and suggested that additionalflattening had occurred associated with the intrusion ofthe magma. He reported northeast vergence in themeso-scale folds crosscut by the intrusive contact. The Peninsula Batholith intruded at 540 ± 4 Ma(Armstrong et al., 1998).

Tygerberg Formation on Robben IslandThe Tygerberg Formation sediments can be locallydivided into three general lithological groups: tan togrey sandstones, interbedded greywacke and siltstone,and fine dark grey slates. These lithologies are similar tothose described by Von Veh (1982) at Sea Point (Figure 1). These rocks form the bedrock of RobbenIsland and outcrop along the rocky coasts and in quarryexcavations. Rare interior outcrops on the island arebadly weathered and were not utilized in this study. A dolerite dyke cuts across the Tygerberg sediments inthe southwest corner of the island. The generaldistribution of these facies is shown on Figure 2, whichalso gives the location of photos in the descriptionswhich follow. These units are unconformably overlainby Quaternary beach and dune sands and calcrete whichcover nearly the whole interior of the island (Theron et al., 1992).

SandstonesThe sandstones of the Tygerberg Formation outcrop onthe east and west-northwestern coasts of Robben Island(Figure 2). They are tan to light grey in colour,corresponding to slight changes in composition,although they tend to discolour into a darker brownwith weathering. In general, the Tygerberg Formationsandstone is medium to coarse grained. Beddingthickness is typically in the range of 10 to 30 cm but

reaches 50 cm locally in the northwest corner of theisland. Orthogonal joints cut the sandstones in threemajor orientations: north to south, east to west andnorthwest to southeast (Figure 3A). Grey-greenreduction stains may be observed on the surfaces ofthese sandstones and are concentrated along the jointsurfaces. Spheroidal weathering between joint surfacesis a characteristic feature of the outcrop appearance(Figure 3A). Where bedding and cleavage are sub-



Figure 2. Generalized lithologic map of the rock type distributions and major structures within the Tygerberg Formation on Robben Island.

Locations of field photos in Figures 3, 4 and 6 are indicated.




61

parallel and cleavage is closely spaced, the sandstonesmay take on a slaty appearance but bedding structuresare still very well preserved. Parallel lamination andtrough cross-bedding are common. Discrete intervals ofconvolute bedding (also reported by Nakashole, 2004)are locally present and sometimes deformed by widely-spaced pressure solution cleavage (Figure 3B). In general, pressure solution cleavage in the sandstonesis weaker and more widely spaced than in the morepelitic units of the Tygerberg Formation. The spacingand orientation of cleavage planes varies with beddingthickness and with slight, often gradational variations ingrain size and clay content (Figure 3C). Sand-filledburrows were reported by Nakashole (2004), suggesting

active bioturbation in these facies, although this studydid not confirm these observations.

The sandstones are similar to Facies A of Von Veh(1982), who interpreted the mechanism of deposition asfluidization, grain flow or high density turbidity currents.The frequency of convolute bedding horizons suggestshigh rates of deposition and burial.

Interbedded shales and greywacke/siltstonesThe siltstones of the Tygerberg Formation are darker incolour than the sandstones and contain metamorphicchlorite and diagenetic or metamorphic pyrite(Mbangula, 2004; Nakashole, 2004). The greywackegrain fraction contains sub-rounded to angular lithic,

Figure 3. Features of sandstone-rich strata. (A) Sub-horizontal bedding surface view of a coarse sandstone bed on the southeast corner of

Robben Island. Three major joint orientations are suggested by white outlines on the lower left of the photo. Spheroidal outcrop surfaces are

created in situ by joint-controlled weathering processes. (B) A thin convolute bed in the tan sandstones, northwest corner of Robben Island.

Brunton compass is 7 cm across casing and indicator points toward North. Bedding contacts indicated by solid white line. Pressure solution

cleavage cuts bedding at a high angle and refracts gently to a slightly steeper orientation within convolute sandstone bed. Modification of

convolute bedding forms qualitatively hint at the degree of shortening along pressure solution cleavage. (C) Cleavage spacing and orientation

changes across lithologic variation in sandstone units near the northwest corner of Robben Island. Pen is 14 cm long and points north.

Bedding cuts horizontally across photo (solid white lines) and cleavage is locally sub-perpendicular, cutting vertically across photo. Changes

in bedding lithology are annotated on the right side of photograph with variation from clay rich toward coarse sand toward the west-

northwest. Corresponding changes in cleavage spacing are annotated on the left side of the photo and slight changes in orientation are

suggested by the dashed white lines. (D) Interbedded shales and greywacke siltstones outcrop beautifully in the axial region of a north-

plunging anticline near the northwest corner of the island.



Figure 4. Features of dark grey slates. (A) Combined ripples in dark grey slates, Jan van Riebeeck Quarry. (B) Rip-up clasts of light-coloured

siltstone in fine dark grey slates, Jan van Riebeeck Quarry. Clasts are weakly oriented with long axis approximately horizontal across

photograph. Median axis is approximately up-down. Clasts are highly flattened perpendicular to cleavage in the third dimension.

(C) Fine-scale planar lamination and convolute slump textures in dark grey slates, Jan van Riebeeck Quarry. Diameter of pen is 12 mm.

(D) Pressure solution cleavage in dark grey slates, Jan van Riebeeck Quarry. Left edge of photo is annotated to show scoured bedding surfaces

(sub horizontal, solid white lines) and pressure solution cleavage (dashed white lines). Gradations within each bed correspond to changes

in orientation and spacing of cleavage creating anastomosing geometries. General regions of closely and broadly spaced cleavage (relatively

speaking) are indicated. Repetitive graded beds result in closely spaced cleavage transitioning toward increasing spacing upsection. Bedding

is upright as shown by the subtle crossbedding. (E) Anastomosing slaty cleavage seen intersecting the plan of the bedding surface, Jan van

Riebeeck Quarry. Pen is 12 mm in diameter and points toward the northwest. Dashed white lines on left side of photo annotate cleavage-

bedding intersection lines showing gently wavy, anastomosing geometry. (F) Stereonet plot of axial planar pressure solution cleavages from

all across Robben Island (n=107). Mean cleavage plane shown, circle around mean pole indicates 95% confidence cone

SOU

TH

AF

RIC

AN

JO

UR

NA

L O

F G

EO

LO

GY

ST

RU

CT

UR

AL G

EO

LO

GY

OF R

OB

BE

N I

SLA

ND

Fig

ure

5.

Str

uct

ura

l m

ap

of t

he

Tyg

erbe

rg F

orm

ati

on o

utc

ropp

ing

on R

obbe

n I

sla

nd

. A

eria

l ph

otog

raph

y a

cqu

ired

in

2008 g

ener

ousl

y pr

ovid

ed b

y C

hie

f D

irec

tora

te:

Na

tion

al

Geo

spa

tia

l In

form

ati

on (

DM

C j

ob

3318C

2008 1

27).

In

set

(A)

Cle

ava

ge f

an

nin

g in

mes

o-sc

ale

cyl

ind

rica

l fo

ld.

Ster

eon

et s

how

s π-

dia

gra

m s

olu

tion

for

fol

d a

xis

from

bed

din

g m

easu

rem

ents

. St

ar

sym

bol

= f

old

axi

s (1

5/1

22).

In

set

(B)

Ad

jace

nt

but

oppo

site

ly p

lun

gin

g fo

lds

sepa

rate

d b

y cr

ypti

c be

dd

ing-

para

llel

wre

nch

su

rfa

ce.

Bed

din

g is

con

tin

uou

s fr

om e

ach

axi

s ou

t to

th

e w

ren

ch f

au

lt,

the

posi

tion

of

wh

ich

is

rest

rict

ed t

o a

1.5

m-w

ide

lim

b-pa

ralle

l zo

ne

wit

h m

inor

bed

din

g d

eflec

tion

an

d n

o n

otic

eabl

e br

ittle

da

ma

ge z

one

on e

ith

er s

ide.

Wre

ck o

f th

e 60 m

Sea

Ch

alle

nge

r fo

r sc

ale

. (C

) St

ereo

net

plo

t of

mes

o-sc

ale

fol

d a

xes

acr

oss

Rob

ben

Isl

an

d (

n=

18).

Ea

ch f

old

axi

s w

as

calc

ula

ted

usi

ng

a m

inim

um

of

4 b

edd

ing

att

itu

des

rep

rese

nti

ng

both

lim

bs,

alth

ough

spa

tia

lly d

eta

iled

da

ta h

as

been

om

itte

d f

rom

th

e m

ap

(Fig

ure

5)

for

cla

rity

. (D

) B

lock

dia

gra

m s

how

s th

e se

nse

of

rota

tion

on

a t

ypic

al

Mod

e II

fa

ult o

n t

he

wes

t co

ast

of

Rob

ben

Isl

an

d.




63

feldspar and quartz grains in a clay-rich matrix.Typically, the siltstones are interbedded withgreywackes and slight cleavage refraction is observedacross lithologic contacts. These siltstone/greywackeintervals are 1 to 20 cm thick, separated by thin shalebeds. Differential erosion of the shale beds is commonalong the coastline, resulting in excellent bedding plane exposures in the axial region of 10’s of metres-scale folds (Figure 3D). These strata are similar to Facies B defined by Von Veh (1982), who interpretedthem as deposited by turbidity currents in a “proximal”setting.

Dark grey slatesDark grey fine-grained slates outcrop on the southern tipand northwestern corner of the island and were quarriedat Jan van Riebeek Quarry in the south and RangatiraQuarry in the north (Figure 2). Delicate sedimentary andsoft-sediment deformation features are common,including combined ripples and bidirectional cross-bedding (Figure 4A), sinusoidal ripple lamination,graded bedding, formsets of ripple-drift crosslaminations, regular cross laminations, light grey rip-upclasts (Figure 4B), and delicate soft-sedimentdeformation structures (e.g. Figure 4C). The slates havegently anastomosing cleavage, which refracts slightlyacross bedding when cutting at a high angle (Figure 4D).The cleavage-bedding intersections as viewed on thebedding surface (Figure 4E) are also anastomosing andvariable in spacing.

This lithology is consistent with Facies D of Von Veh(1982), who also noted very fine laminations alternatingbetween dark (clay rich) and light (silt rich) as well asthe delicate sedimentary structures characteristic of fluidescape processes.

Dolerite dykeA single dolerite dyke occurs on Robben Island. The dyke is similar in lithology and orientation to thelocal Cretaceous (132 ± 5 Ma) False Bay dyke swarm(Reid et al., 1991). The dyke is tabular and trends east-southeast to north-northwest across the south coast ofthe island, crosscutting bedding and cleavage in theTygerberg Formation. It is preferentially eroded and the contacts are not exposed. Although the dolerite doesnot outcrop in situ, it is abundant in boulders andcobbles within a tabular zone approximately 20 m thick(Figure 2).

SummaryThe sedimentary units on Robben Island are dominatedby turbiditic sequences (similar to the Sea Pointexposures, Von Veh (1982)) and all the representativelithologies show evidence of rhythmic normal grading.Coarse sediments dominate most sections and thesediments are chemically immature. Sole marks and ripup clasts (Figure 4B) testify to high rates of sedimentflow and deposition. The rate of deposition oftenexceeded the rate of static dewatering, as demonstrated

by symmetric dewatering structures such as convolutebedding, load casts and flame structures which arepresent in all facies. Fine horizontal laminations are alsocommon in all facies (Figure 4D.) Locally, bidirectionalripples and cross-bedding suggest multi-directionalcurrents were active during deposition. Massive muddy olistostromal beds were observed at Sea Point(Von Veh, 1982), and ~8 km north of Robben Island at Silwerstroomstrand (C. Rowe et al., unpublisheddata).

Conclusive evidence of bioturbation is rare, althoughsand-filled burrows were described by Nakashole(2004). The lack of fossils in the Tygerberg Formationlimits the precision of palaeo-environmentalinterpretations. In the modern world, shallow marineenvironments are characterised by intense bioturbationand biological reworking. Nakashole (2004) relied onthe rarity of bioturbation fabrics to prefer deep waterturbidite origin over a shallower water tempestite orcontourite setting for the Tygerberg Formation.However, an age constraint on the Tygerberg Formationis given by the intrusion of the Cape Granite Suite. The nearest intrusion to Robben Island is the PeninsulaPluton cropping out at Sea Point (540 ± 4 Ma, Armstronget al. (1998)). Globally, the widespread appearance ofburrowing behaviours which contribute to bioturbationis known as the “substrate revolution” (Bottjer et al.,2000). This stratigraphic event is recorded in Cambrianshallow marine sediments worldwide. Therefore, thepaucity of bioturbation in the Tygerberg Formation isnot considered an indication of deep water, and thedepth remains unconstrained.

Structural featuresThe structural features exposed at Robben Island aredominated by: strong vertical northwest-southeastpressure solution cleavage, asymmetric meso-scale folds,cryptic bedding-parallel shear zones (sometimes bearingquartz-cemented breccia), and east-west vertical jointsand quartz veins (Figure 5).

Pressure Solution CleavageAll outcrops of Tygerberg Formation on Robben Islandare crosscut by strong pressure solution cleavage. The cleavage in the clay-rich lithologies is planar andparallel on a regional scale (Figure 4F) but in sandierlithologies the cleavage shows distinct local fanningaround fold axes (Figure 5A). As described above, thepressure solution cleavage is locally anastomosing andrefracts gently through graded beds. In general, cleavageis axial planar to the major folds and bedding parallelbetween fold axes. On the scale of the whole island, thecleavage is very well aligned (Figure 4F) with the meancleavage plane oriented 330/86NE (axial planar to theupright folds.)

FoldsThe structure of Robben Island is dominated bynorthwesterly and southeasterly plunging folds, which

often occur in anticline-syncline pairs (Figure 5). The half-wavelength of the folds is typically 40 ± 10m.Axial planes are upright and parallel to the regionalpressure solution cleavage. The shape of fold hingesvaries, with concentric folds occurring in sandiersections (e.g. Figure 3D) and increasingly similar foldmorphologies observed in shalier units (e.g. Figure 5A).Asymmetry of anticline-syncline pairs generally suggestsnortheast vergence. Parasitic folding at scales below themeso-scale folds was only observed in one location. The folds are open to closed with interlimb anglesranging between 35 to 75°.

The eighteen mapped meso-scale folds on RobbenIsland are coaxial with the regional tectonic fabrics andplunge 5 to 20° in either direction (Figure 5C). The foldaxes cluster in either orientation but do not define acontinuous distribution in plunge angle. Folds arecylindrical and doubly-plunging folds were not observedon the island. This could be attributed to limited along-trend exposure, however, one would expect thatrandom exposure of doubly plunging folds wouldproduce an array of fold axes around a central meanvalue. Our data clearly form two discrete clusters at 10 to 15° plunge. The clusters trend in oppositedirections although there is considerable scatter in trenddirection. It was not possible to trace fold axes acrossRobben Island where they might be covered byHolocene sediments, suggesting that axial traces maynot be continuous on the km-scale. This distinctivepattern will be further discussed below.

In more than one location, oppositely plunging foldsare closely associated, apparently sharing roughlyparallel limbs (e.g. Figure 5B). The transition zonebetween the two anticline axes in Figure 5B does notshow a gradual rotation of bedding as would beexpected if a syncline separated the anticlinal hinges.Limb orientation is maintained at relatively constantcurvature from both folds, up to a zone approximately1.5m wide where bedding is not exposed. Theseobservations suggest that the unexposed zonerepresents a cryptic wrench fault along beddingseparating the shared limbs. A similar scissor is observedfarther south on the west coast of the island (Figure 5D.)

Faults and shear zones, quartz-cemented brecciasThe majority of faults are inferred from poorly exposednarrow zones across which bedding and/or cleavagechange orientation sharply (e.g. Figure 5B describedabove). Although direct indications of shear sense arerare, the bedding orientation discontinuities requiresignificant rotation along the faults. A few small faultsshowed quartz slickenfibres but in most cases thedirection of shear is inferred from drag folding ofbedding. In all cases these indicate oblique slip.

Rarely, quartz-cemented breccias are found in thesezones (Figure 6A), most commonly in sandstone unitsalthough they are inferred in other lithologies by thepresence of quartz-cemented breccia boulders observedin float. The breccias have a chaotic fabric and quartz

veins around them are deformed (flattened andboudinaged) along the pressure solution cleavage,suggesting that these breccias are pre-to syn-shortening.Stretching direction recorded by the boudinage of theseearly quartz veins is parallel to the strike of the pressuresolution cleavage.

At the southern end of the island, an intenselydeformed breccia zone cemented by quartz indicates thepresence of a brittle shear zone (Figure 6B). A tightanticline is asymmetrical toward the southwest, suggestingsouthwest vergence. This anticline structurally overliesan intensely veined shear zone and is interpreted as adrag fold in the hanging wall of a southwest-vergent,bedding-parallel thrust fault. This unique exposuresuggests an association between the early quartz-cemented breccias and shortening by brittle faulting.

Joints and fibrous/blocky extensional quartz veinsQuartz veins were observed throughout the strata atRobben Island and crosscut the pressure solutioncleavage and folds (Figure 6C.) Veins are concentratedin the sandstone-dominated units and do not have anyapparent association with fold hinges or faults. Theygenerally strike east-west and dip steeply south,although there is quite a bit of scatter in the orientations(Figure 6D). The veins are generally blocky but locallyaligned euhedral fibrous crystals are present. Theyplunge gently to the north-northwest, suggesting gradualoblique opening for the veins (Figure 6F). The veins areroughly parallel to the tabular Cretaceous-age False Baydykes (only one on Robben Island; Figure 2).

These veins are therefore kinematically consistentwith, and most likely associated with, Cretaceousextensional rifting coincident with the intrusion of theFalse Bay dykes (Reid et al., 1991).

Environment of deposition and deformationThe depositional environment of the TygerbergFormation is not well constrained by previous studies,although two general themes exist in the literature:Rozendaal et al. (1999) suggest a deepeningocean/continental margin setting, and Von Veh (1982)suggests intercalation of proximal and distal distributaryfans. The significant volume of turbidites implies adynamically deepening depocentre, while thesedimentary structures imply high energy.

One possible match to these environmentalconditions, and the stratigraphic relationship to the moredeformed Swartland terrane, is the setting of a forearc ortrench slope basin overlying an active subduction zone.Depositional basins often develop in the forearc regionof a subduction margin due to high sediment supplyfrom the nearby volcanic arc, and the topographicbasins created by the intricate processes of compressionand extension in accretionary wedges. These may beseparated or continuous with depositional basins on thetrench slope, which share many characteristics withforearc basins but may be distinguishable by theirdeeper water and steeper slope, expressed in finer






65

overall grain size and increased relative importance ofgravity slides. Both basin geometries are predicted byCoulomb models for wedge evolution, especiallymodels incorporating erosion (where appropriate) andthermo-mechanical evolution of the wedge sediments(e.g. Fuller et al., 2006; Kimura et al., 2007).

Lithofacies of forearc and trench slope basinsSeveral recent and modern examples of forearc andtrench slope basins are presented for comparison to the Tygerberg Formation. Our examples range from thelandward edge of the forearc basin (the Pliocene

Quinault Formation, Cascadia (Campbell et al., 2006)),broad forearc settings (the modern Kumano Basin, Japan(Park et al., 2002; Moore et al., 2009) and the CretaceousGreat Valley Sequence, California (Ingersoll, 1979)),outer forearc basin (the Mio-Pliocene Awa Group, Japan(Tokuhashi, 1989)), trench slope (Cretaceous MatanuskaFormation, Alaska, USA, (Trop, 2008), and trench slope-trench (the Eocene Sitkalidak Formation, Alaska, USA(Moore and Allwardt, 1980)). All the examples aredominated by turbidite fans which overlap laterally andthrough vertical section. They are all relatively long,narrow basins which unconformably overlie higher

Figure 6. Occurrence of quartz veins and breccia cements on Robben Island. (A) Intrafolial quartz-cemented breccia extended along

pressure solution fabric. Length of pen = 14cm. (B) Hanging wall asymmetric fold pair adjacent to small southwesterly-vergent quartz-

cemented brecciated thrust zone. (C) Late transtensional quartz veins crosscut pressure solution fabrics. Aligned fibres and C-axes of

subhedral crystals show direction of vein opening (black arrows), which is consistent across different vein orientations (D) Stereonet plot

of late transtensional tabular quartz veins from all across Robben Island (n=37). Extension direction as recorded by orientation of quartz

fibres shown with black squares.

grade, more deformed meta-greywackes and shales,comprising accretionary wedges. They share thecharacteristic paleocurrent patterns with two maximaparallel and perpendicular to the basin axis,representing sediment flow into the basin and along theaxis. Large-scale slumping and slope collapse depositsare prominent in all the examples, particularly in thetrench slope settings. Lithologically, coarse sands andconglomerates dominate the landward forearc basin(Campbell et al., 2006), transitioning to finer silts andshale in the trench slope to trench deposits (e.g. Mooreand Allwardt, 1980; Trop, 2008). Some of the basinscontain regionally continuous tuff beds: airfall depositsfrom the nearby arc (Great Valley, Kumano: Ingersoll,1979; Moore et al., 2009), while the others do not butmay have a significant volcaniclastic component in thesandstones. Water depth indicators in the mid-forearcbasin, where they are reported, are generally bathyal(~200 to 2000m) (Awa Group, Great Valley, Kumano:Tokuhashi, 1989; Ingersoll, 1979; Park et al., 2002).Several examples show evidence of tectonic deepeningof the basin by localized normal faulting either pre- orsyn-depositionally (Kumano, Awa, Matanuska: Park et al., 2002; Tokuhashi, 1989; Trop, 2008).

In terms of the basin geometry, the Kumano Basinsediments onlap the irregular topography created by theextension of the wedge landward of the outer rise of the accretionary wedge, and even lap seaward over theouter rise into the trench slope basin (Moore et al.,2009). Moore and Allwardt (1980) distinguished twostructural zones in the Sitkalidak Formation: the seawardedge of the basin was sheared and then horizontallyshortened, while the landward part of the basin showsupright folds and axial planar cleavage, with no evidenceof shearing or substantial intraformational faulting.

Comparison to the Tygerberg FormationSeveral of the diagnostic features indicating relativelyhigh-energy, episodic deposition for the forearc andtrench slope basins described above were also observedon Robben Island and reported elsewhere in theTygerberg Formation. The dominance of turbidites andoccurrence of olistostromes and slump features suggestdeposition by periodic high-volume events. The combined ripples (Figure 4A) require multi-directional flow, the rip-up clasts (Figure 4B) suggesttransient high-velocity flow conditions, and theabundant convolute bedding (Figure 3B) anddewatering structures (Figure 4C) present in all faciessuggest high rates of deposition. Hummocky crossstratification in massive sandstone beds was notobserved on Robben Island, but has been recorded inthe Tygerberg Formation at Sea Point (Von Veh, 1982)and at Grotto Bay about 10 km north (C. Rowe et al.,unpublished data). Scoured bedding contacts were notobserved in the Tygerberg Formation on Robben Islandalthough they have been reported elsewhere (Von Veh,1982). Von Veh (1982) also reported varying butgenerally bimodal current directions for the Tygerberg

Formation turbidites at the Sea Point Contact, consistentwith the other examples.

Based on the multi-directional current indicators,frequent coarse sands, rip-up clasts and abundantdewatering structures, we concur with the interpretationby previous researchers that the Tygerberg Formationwas deposited in a deepening basin (Rozendaal et al.,1999) by migrating distributary channels on coalescingfans (Von Veh, 1982). Although there is yet insufficientdata to uniquely determine the depositional setting ofthe Tygerberg Formation, we note the similarities toforearc and trench slope basins in which large volumesof chemically immature marine sediment are deposited,dominantly as turbidites building distributary fans, withtectonic basin deepening acting to generally maintainconsistent water depth. As pointed out by Cawood et al.(2009), forearc basins are dominated by lithicgreywacke-shales as the primary source is an arc, butbackarc basins are more quartzose as they also receivesediment from the landward continent. In addition, theclosure of a back-arc basin is associated with highgeotherms while the closure of a fore-arc basin remainscold (Cawood et al., 2009). The unconformablerelationship to the Swartland terrane described byBelcher and Kisters (2003) and Von Veh (1982) isconsistent with the relationship between a flat-lyingforearc basin and the underthrusted sediments of theaccretionary wedge.

Implications of fold orientation distributionModels of folding in transpressional systems predictparallel fold axes which rotate with material lines towardthe shear direction during increasing strain. Thisparadigm treats fold axes as features which, onceformed, rotate as passive markers during increasingshear. These folds are typically shown as symmetricaland upright horizontal or gently doubly-plunging, withparallel trends (e.g. Davis and Reynolds, 1996;Sanderson and Marchini, 1984; Krantz, 1995, andreferences therein). In this paradigm, the long axis of thestrain ellipse within the shear zone rotates from foldinitiation at approximately 30° to the shear zoneboundaries (perpendicular to maximum compressivestress) to smaller angles as it extends with increasingoblique strain. A prediction of this model is that theangle between fold axes and the direction of sheardepends on both the angle of obliquity and increasingstrain.

A contrasting paradigm is provided by the exper-iments of Tikoff and Peterson (1998) and theobservations of Titus et al. (2007). Tikoff and Peterson(1998) performed experiments using sheets of plasticineand silicon to investigate the fold morphologies andorientations formed during transpression. They variedthe angle of the shear zone with respect to the directionof maximum compression (� = 0° when compression isparallel to the margin; � = 90° when compression isnormal to the margin). Their results indicate that foldaxes initiate with plunges around 5 to 10° even while






67

compression and shear are horizontal, and thatsignificant scatter in trend of the fold axes does notdecrease with increasing strain. Most importantly, theyshowed that fold axes do not follow material lines withinthe shear zone – therefore the fold axis orientation isdecoupled from both angle of obliquity and the totalstrain, but would be expected to steepen with increasingstrain if rocks behaved passively as the system evolved.In essence, these experiments show that fold axes arenot maintained as passive markers, but rather that theycan continue to move relative to bedding surfaces whichroll over the fold axes as the folds tighten (Smallshire,1999).

Our fold axis observations (Figure 5C) show a similarpattern to that produced by Tikoff and Peterson (1998).Two distinct clusters of fold axes are shown. The northwest plunging folds trend 310 ± 14° andplunge 13 ± 3°. The southeast plunging folds trend 136 ± 14° and plunge 13 ± 3°. The divergent plungesand consistently variable trends of fold axes areconsistent with transpressional strain.

Titus et al. (2007)’s structural and paleomagneticstudy of en echelon folds formed around a slightlytranspressive segment of the San Andreas Fault in centralCalifornia confirmed the patterns of fold axis trendproduced by Tikoff and Peterson (1998). Further, theirpaleomag results require vertical axis rotations of thefolded strata in addition to the horizontal axis rotationimposed by tightening folds. The juxtaposition of thesetwo rotations suggests complex rotational shearing isnecessary along limbs of folds in order to accommodatethe necessary stretching of limbs. Scissor faults such asthose shown in Figures 5B and 5D may have formedduring similar intrafolial rotations.

Shortening recorded by the structures on RobbenIslandThe pervasive northwesterly to southeasterly strikingvertical pressure solution cleavage on Robben Islanddemonstrates strong northeasterly to southwesterlyshortening. Spread in cleavage orientation (Figure 4F)reflects anastomosing (Figure 4E) and fanning (Figure 5A) of cleavage planes, suggesting the cleavageformation was early to syn- folding. The foldasymmetries suggest predominantly northeast-vergencewith local southwest-vergent folds and faults (e.g. Figure6B) representing antithetic structures. These arepreferentially associated with zones of quartz-cementedbreccias, suggesting they represent zones of fluidadvection and hydrofracture. Although strain markersare scarce, where the cleavage is observed crosscuttingsedimentary structures it is possible to comment on the bulk shortening achieved by pressure solution. In the sandstone units where cleavage crosscutsconvolute bedding (e.g. Figure 3B) we estimateshortening on order 10 to 20%. In clay rich lithologieswhere pressure solution is more efficient, the volumeloss, and therefore total shortening, is much greater. In Jan van Riebeeck Quarry, the rip-up clasts shown in

Figure 4B are moderately well-aligned. The large clastsare extremely thin along an axis normal to the pressuresolution cleavage (for example, the large clast in Figure 4B is 15 cm: 5 cm : 0.2 cm; sampling limitationsin the World Heritage Site prevented quantitativeinvestigations of a statistically relevant number of clasts).On this basis, we estimate that pressure-solutionshortening in the clay-rich lithologies may be severaltimes greater than the shortening in the sandstones. As demonstrated by Tikoff and Peterson (1998)’sexperiments, cross-section balancing acrosstranspressional folds is likely to overestimate the totalshortening as it does not account for out-of-planeextrusion. Strain ellipses in transpression show flattened“pancake” shapes (Fossen and Tikoff, 1993), regardlessof strain partitioning or the local mechanisms of strainaccommodation. Analog models have predicted atransition from strike-slip dominated, sub-verticalstructures toward thrust-dominated, moderately dippingstructures at oblique convergence angles in the range of � = 15 to 25° (Casas et al., 2001). Given the sub-vertical orientation of faults and folds observed onRobben Island, and the wrench and rotational faultmotions, we suggest that the Saldanian margin duringlate-Neoproterozoic to early Cambrian was oriented at � = 15 to 25° or less, relative to the direction of plate convergence. Simply stated, the direction ofmaximum shortening on Robben Island, as in otherexamples of transpressional deformation, does notcorrelate to the direction of maximum principlepalaeostress.

The numerical models of Merle (1997) demonstratethat the orientation of the plane of principle flattening(represented on Robben Island by the axial planarpressure solution cleavage, Figure 4F) does not vary inorientation when the angle of oblique convergencevaries. Merle (1997) points out that the unusualsymmetry of the finite strain ellipse produced duringtranspression renders finite strain analyses relativelyuseless in interpreting strain paths during transpression.Similarly, the experiments of Tikoff and Peterson (1998)and the fold axis orientations observed by Titus et al.(2007) are symmetrical and do not distinguish betweensinistral and dextral transpression.

As such, we are unable to measure the exactshortening represented by the structures on RobbenIsland, or determine the sense of transverse motion(whether sinistral or dextral). To differentiate, we wouldneed specific measurements of the rotation in plungedirection of the long axis of the finite strain ellipse(Merle, 1997). Unfortunately, no conclusive strainmarkers recording this information were observedduring our field studies. Citing Kisters et al. (2002)’sobservations of dominantly sinistral motion for theColenso Fault (followed by a very brief episode ofdextral slip at about 540 Ma), and the southwardpropagation of collision indicated in the Kaoko andGariep Belts, sinistral shear sense is considered morelikely.

Regional tectonic eventsNumerous orogenic belts dividing older crustal terranesformed during the construction of southwesternGondwana in the late Neoproterozoic – early Cambrian.The closure of the Adamastor Ocean, uniting Africanand South American blocks, was a complex, multiphaseevent. Many crustal blocks were assembled (Rapela et al., 2007; da Silva et al., 2005), probably involvingrotations in plate motions including changes insubduction polarity as seen in similar modern settings(Cawood et al., 2009). The southward-youngingorogenic belts along the western margin of southernAfrica are enigmatic because structural studies suggesthighly oblique strain, with the consequence of lateraldisplacement of terranes, and the geometry ofsubduction and collision is often unclear.

The Saldania Belt, at the southern end of this trend,introduces particular questions since although there isabundant evidence of crustal shortening, there is nodirect evidence of continental collision or an associatedvolcanic arc that would indicate subduction polarity.Therefore, the tectonic environment of deformation hasnot been adequately explained.

The Kaoko Belt (northern Namibian coast) recordsoblique accretion of an exotic terrane (the CoastalTerrane, during east-dipping subduction; (Goscombeand Gray, (2007), or west-dipping subduction at 650 to 600 Ma, with the Granite Zone of the DomFeliciano Belt as the associated arc complex;(Oyhantçabal et al. (2009)) and compression beginningat 660 Ma, with peak metamorphism at 585 to 560 Maand shear zones active until 530 Ma (Goscombe andGray, 2008).

In the Gariep Belt (southern Namibia-South Africa),basin filling and shortening is dated at 576 to 573 Ma and deformation peaked between 545 to 530 Ma(Frimmel and Frank, 1998; Frimmel et al., 1996; Frimmeland Basei, 2006). In this case, the basin is interpreted asa retro-arc extensional setting during west-dippingsubduction, again citing the Dom Filiciano granites asthe associated arc (Frimmel and Basei, 2006; Basei et al.,2005). Meanwhile, de Ameida et al. (2002) interpret theCamaquã Basin (and equivalents) in southern Brazil toalso represent retro-arc spreading during the same west-dipping subduction. These two possibilities aregeometrically incompatible with one another based oncurrent plate reconstructions.

The relationship between the Gariep Belt and theSaldania Belt is essentially unknown, although someauthors have suggested lateral equivalency based on theolder Neoproterozoic correlations between Gariep rocks and the Cango Inlier (e.g. Frimmel et al., 1996) or apparent strike-alignment and similar age (e.g. Dunlevey, 1992). More recent reconstructions alignthe Saldania Belt with the ~550 Ma Ross Orogen inAntarctica (e.g. da Silva et al., 2005), separated alongstrike but more consistent with the orientation of theKalahari craton in some reconstructions (e.g. Gray et al.,2008).

In the Saldania Belt, oblique shortening wascomplete by the intrusion of the undeformed units in theCape Granite Suite (e.g. Peninsula Batholith at 540 ±4 Ma; Armstrong et al. (1998)), simultaneous with thelate-stage reversal in transpression direction at about 540Ma (Kisters et al., 2002). In contrast to the Kaoko andGariep Belts, no collisional phase or metamorphic eventfollowed. As pointed out by Frimmel and Frank (1998),this southward-younging tectonism is consistent with a zipper-like, south-progressing closure of theAdamastor Ocean, although other models are possible (e.g. southward extrusion of the arc complex, asdiscussed by Miller et al. (2009)).

The youngest detrital zircons reported in theTygerberg Formation are ~560 Ma, (uncertainty notreported; Armstrong et al., 1998), roughly the same ageas the oldest Cape Granites (552 ± 6 Ma; Scheepers andArmstrong, 2002). The Cape Granites are the onlyknown source for zircons of this age. Folds and cleavagein the Tygerberg Formation are crosscut by the late-tectonic Sea Point granites (540 ± 4 Ma; Armstrong et al.,1998; Von Veh, 1982), which solidified at 3 to 4 kbar(Villaros et al., 2006) (approximately 11 to 15 km depthassuming a typical psammitic crustal density of 2650 kg/m3). Volcanics erupted at the same structurallevel as the granites (following their exhumation) by 515 Ma (Scheepers and Poujol, 2002). These constraintsallow a minimum estimate of the regional uplift rateduring the early Cambrian, between the intrusion of theSea Point granite and the eruption of the PostbergIgnimbrite: 11 to 15 km / ~25 Ma gives an uplift rate of~0.4 to 0.6 mm/yr. This minimum estimate of uplift rateis less than or similar to uplift rates in currently activeareas of transpression (e.g. Southern Alps at 6 to 9 mm/yr; Little et al., 2005, 0.13 - 0.35 mm/yr SanAndreas fault in the Santa Cruz Mountains; Valensise andWard 1991).

This uplift was associated with a brief period of localextension by the formation of elongate grabens parallelto regional structure, (the intra-orogen pull-apart basins suggested by Rozendaal et al. (1999)) followed bylong term regional subsidence (Tankard et al., 2009).This fault-parallel, graben-bounded geometry is wellpreserved today (Figure 1) and the conglomerates areunconformably overlain by the regionally extensivePaleozoic Table Mountain Group (Hartnady et al., 1974;Theron et al., 1992; Tankard et al., 2009). Although someprevious authors have suggested that the rifting wasregionally significant and the extent of theconglomerates was originally much greater than is seentoday (e.g. Theron et al., 1992), no direct evidence foreither voluminous deposition or significant normalfaulting at this time has been reported.

Two tectonic settings are suggested, which couldexplain the deposition and deformation of the TygerbergFormation in a supra-subduction setting, link theregional uplift to the switch in shear sense on the Colenso Fault at 540 Ma, and provide context for thedeposition of the graben-bounded conglomerates.






69

Triple-junction migration (or ridge-trench interaction) isassociated with transient localized extension caused bythe passage of a thermal high with or without associatedslab window volcanism (Unruh et al., 2007; Farris andPaterson, 2009; Ayuso et al., 2009; Cole and Steward,2009), which we suggest may correlate to the CapeGranite Suite. Alternatively, migration of releasing bendsalong a transform boundary can produce chains ofgrabens along strike-slip faults without regionalextension (Seeber et al., 2010; Wakabayashi et al., 2004).These two models can be compared and tested byfurther studies distinguishing: the age and distribution ofthe lower Cambrian conglomerates, and the ageprogression and tectonic setting of melting whichproduced the Cape Granite Suite, in particular, the relationships in age and petrogenesis between the S-type and I-type granites.

Although numerous authors have suggested asubduction setting for the Saldania Belt and associatedgranites (Belcher, 2003; Da Silva et al., 2000; Dunlevey,1992; Scheepers, 1995; Stevens et al., 2007 and others),no associated arc complex has been uniquely identified.Our data presented here do not allow us to suggest adirection of subduction vergence, only that thesedimentological and tectonic setting is consistent with asupra-subduction zone basin, as would be expectedaround the edges of the disappearing Adamastor Oceanbasin. If the closure of the Adamaster Ocean was acomplex, multi-stage event involving multiple terranesas implied by Germs and Gresse (1991) and Stanistreetet al. (1991), a forearc basin may display complexprovenance and source directions (e.g. the WrangelMountains Basin, which laps onto the accretedWrangellia Terrane of southern Alaska, Trop et al.(1999)).

In similar transpressional subduction environments,significant strike-slip in the forearc region has beenshown to laterally separate arc complexes fromassociated forearc basins (e.g. Alaska’s Kodiak Complex,Sample and Reid, 2003). In addition, periods ofmagmatic hiatus are reported in numerous Mesozoicsubduction complexes around the Pacific Rim (e.g. Cretaceous Mexican Trench, (Solari et al., 2007) orMiocene Cascadia (Dostal et al., 2008)) during whichsignificant transverse motion may have occurred.Therefore, the highly oblique transpression in theTygerberg Formation implies the potential for substantialforearc displacement. We note that due to the obliquesetting, matching of ~750 Ma belts likely does not applyto correlations of ~550 Ma belts, especially whenattempted on a Mesozoic (pre-Gondwana rift) tectonicmap, as demonstrated by the spatial mismatch ofmetamorphic and magmatic ages (Gray et al., 2008,Figure 12).

As obliquity is characteristic of all the Neoproterozoicbelts on the western margin of southern Africa, it isunlikely that the current attempts to reconstruct theCambrian geometry of southwest Gondwana can beaccurate when using a plate fit model based on the mid-

Mesozoic configuration. A large number of the attemptsat reconstruction use these Jurassic plate configurations(e.g. Bossi and Goucher (2004); Da Silva et al. (2005);Frimmel et al. (1996); Frimmel and Frank (1998);Gaucher et al. (2004); Gresse (1995); Gresse andScheepers (1993) and Rozendaal et al. (1999) which allrefer to Porada (1979; 1989), which reuse the models ofSmith and Hallam (1970) and Norton and Sclater,(1979)). Therefore, future reconstructions of Cambrian(late Pan-African) plate geometries must consider thisadditional complication.

ConclusionThe Tygerberg Formation was deposited in a high-energy, high deposition-rate, tectonically deepeningbasin during the late Neoproterozoic to very earlyCambrian. These characteristics as well as theunconformable relationship to highly sheared, slightlyhigher metamorphic grade underlying sedimentary unitsare consistent with deposition in a forearc basin. Syn- toimmediately post-depositional deformation occurred inan overall regime of highly oblique transpression andrapid regional exhumation ended the life of the basin.Although direct correlations with South Americanequivalents are not yet possible and most attempts atregional reconstruction end at the Gariep Belt, we hopethis study might contribute an additional constrainttoward filling the gaps in Pan-African southernGondwana.

AcknowledgementsIn order to produce a high-density structural data set ina short time period and provide for student training andresearch experience, we formed “Team TygerbergTerrane”, an informal group at the University of CapeTown Department of Geological Sciences. Students frompre-Honours to MSc level joined the field team andparticipated in preparation of this manuscript. Fundingfor our Robben Island field work, as well as furtherexpeditions of Team Tygerberg Terrane, was generouslyprovided by the Geological Society of South Africa: LE Kent Trust Fund, and we thank Craig Smith forsupporting this unconventional project. We thank StevenJansen and Sue Kirschner from the Chief Directorate:National Geospatial Information, South Africa forpermission to use the base image for the structural map(Figure 5). We are very grateful to Richard Sherley(Avian Demography Unit, UCT) and Mario Leshoro(Ecologist, Robben Island Museum) for their time andsupport for our initial reconaissance work. Rabia Damonof the Robben Island Museum made our field workpossible and enabled the project to go forward. RichardWhiting facilitated our field studies on the island bypreventing conflicts with seabird breeding activities onRobben Island. Professor Alex Kisters, Dr John Rogersand Dr Francesca Meneghini provided valuable insightson the regional tectonic history and structural modelswhich informed the conclusions of this work. Dr Federico Farina, Professor Ian Buick and Dr Arnaud

Villaros are gratefully acknowledged for conversationson the Cape Granites. Stereonets were plotted usingStereonet v. 6.3 X for Macintosh © 2005, and ProfessorRichard Allmendinger is sincerely thanked for makingthis software freely available for non-commercial use.We thank Professor’s Alex Kisters and Udo Zimmermannand Editor Jay Barton for constructive feedback whichgreatly improved this manuscript.

Author contributionsRowe wrote the manuscript with contributions from vanRensburg. Backeberg compiled the maps withcontributions from Maclennan. All authors contributed tofield data collection and observations and completion ofthe manuscript.

References Armstrong, R., de Wit, M. J., Reid, D., York, D. and Zartman, R. 1998. Cape

Town’s Table Mountain reveals rapid Pan-African uplift of its basement

rocks. In: Gondwana 10: Event Stratigraphy of Gondwana, Journal of

African Earth Sciences. 27, 1A. pp. 10–11.

Ayuso, R.A., Haeussler, P.J., Bradley, D.C., Farris, D.W., Foley, N.K. and

Wandless, G.A. 2009. The role of ridge subduction in determining the

geochemistry and Nd-Sr-Pb isotopic evolution of the Kodiak batholith in

southern Alaska. Tectonophysics, 464, 137–163.

Basei, M.A.S., Frimmel, H.E., Nutman, A.P., Preciozzi, F. and Jacob, J. 2005.

A connection between the Neoproterozoic Dom Feliciano

(Brazil/Uruguay) and Gariep (Namibia/South Africa) orogenic belts –

evidence from a reconnaissance provenance study. Precambrian

Research, 139, 195–221.

Belcher, R.W. 2003. Tectonostratigraphic evolution of the Swartland region

and aspects of orogenic lode-gold mineralisation in the Pan-African

Saldania Belt, Western Cape, South Africa. Unpublished Ph.D. thesis,

University of Stellenbosch, South Africa, 302pp.

Belcher, R.W., Kisters, A.F.M. 2003. Lithostratigraphic correlations in the

western branch of the Pan-African Saldania belt, South Africa: the

Malmesbury Group revisited. South African Journal of Geology,

106, 327–342.

Bossi, J. and Goucher, C. 2004. The Cuchilla Dionisio Terrane, Uruguay: An

allocthonous block accreted in the Cambrian to SW-Gondwana.

Gondwana Research, 7, 661–647.

Bottjer, D.J., Hagadorn, J.W. and Dornbos, S.Q. 2000. The Cambrian substrate

revolution. GSA Today, 10 (9), 6.

Campbell, K.A., Nesbitt, E.A. and Bourgeois, J. 2006. Signatures of storms,

oceanic floods and forearc tectonism in marine shelf strata of the Quinault

Formation (Pliocene), Washington, USA. Sedimentology, 53, 945–969,

doi: 10.1111/j.1365-3091.2006.00788.x.

Casas, A.M., Gapais, D., Nalpas, T., Besnard, K. and Román-Berdiel, T. 2001.

Analog models of transpressive systems. Journal of Structural Geology, 23,

733–743.

Cawood, P.A. and Buchan, C. 2007. Linking accretionary orogenesis with

supercontinent assembly. Earth-Science Reviews, 82, 217–256.

Cawood, P.A., Kröner, A., Collins, W.J., Kusky, T.M., Mooney, W.D. and

Windley, B.F. 2009. Accretionary orogens through Earth history. In: P.A.

Cawood and A. Kröner (Editors), Earth Accretionary Systems in Space and

Time. The Geological Society, London, Special Publication, 318, 1–36.

Cole, R.B. and Steward, B.W. 2009. Continental margin volcanism at sites of

spreading ridge subduction: Examples from southern Alaska and western

California. Tectonophysics, 464, 118–136.

Da Silva, L.C., Gresse, P.G., Scheepers, R., McNaughton, N.J., Hartmann,

L.A. and Fletcher, I. 2000. U-Pb SHRIMP and Sm-Nd age constraints on the

timing and sources of the Pan-African Cape Granite Suite, South Africa.

Journal of African Earth Sciences, 30, 795–815.

Da Silva, L.C., McNaughton, N.J., Armstrong, R., Hartmann, L.A. and Fletcher,

I.R. 2005. The Neoproterozoic Mantiqueira Province and its African

connections: a zircon-based U-Pb geochronologic subdivision for the

Brasiliano/Pan-African systems of orogens. Precambrian Research,

136, 203–240.

Davis, G.H. and Reynolds, S.J. 1996. Structural Geology of Rocks and

Regions. John Wiley and Sons, Inc., USA, 776pp.

De Almeida, D.P.M., Zerfass, H., Basei, M.A., Petry, K. and Gomes, C.H. 2002.

The Acampamento Velho Formation, a lower Cambrian bimodal volcanic

package: Geochemical and stratigraphic studies from the Cerro do Bugio,

Perau and Serra de Santa Bárbara (Caçapava do Sul, Rio Grande do Sul,

RS-Brazil). Gondwana Research, 5, 721–733.

Dostal, J., Keppie, J.D., Church, B.N., Reynolds, P.H. and Reid, C.R. 2008.

The Eocene–Oligocene magmatic hiatus in the south-central Canadian

Cordillera: a capture of the Kula Plate by the Pacific Plate? Canadian

Journal of Earth Sciences, 45, 69–82.

Dunlevey, J.N. 1992. Pan-African crustal evolution of south-western Africa.

Journal of African Earth Sciences, 15, 207–216.

Farris, D.W. and Paterson, S.R. 2009. Subduction of a segmented ridge along

a curved continental margin: Variations between the western and eastern

Sanak–Baranof belt, southern Alaska. Tectonophysics, 464, 100–117.

Fossen, H. and 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.

Frimmel, H.E. and Basei, M.A.S. 2006. Tracking down the Neoproterozoic

connection between southern Africa and South America – A revised

geodynamic model for SW-Gondwanan amalgamation. In. C. Gaucher,

J. Bossi (Editors). V South American Symposium on Isotope Geology,

Punta del Este 24–27 August. ISBN 9974-0-0327-X.

Frimmel, H.E. and Fölling, P.G. 2004. Late Vendian closure of the Adamaster

Ocean: Timing of tectonic inversion and syn-orogenic sedimentation in

the Gariep Basin. Gondwana Research, 7, 685–699.

Frimmel, H.E., Fölling, P.G. and Diamond, R. 2001. Metamorphism of the

Permo-Triassic Cape Fold Belt and its basement, South Africa. Mineralogy

and Petrology, 73, 325–346.

Frimmel, H.E. and Frank, W. 1998. Neoproterozoic tectono-thermal evolution

of the Gariep Belt and its basement, Namibia/South Africa. Precambrian

Research, 90, 1–28.

Frimmel, H.E., Hartnady, C.J.H. and Koller, F. 1996. Geochemistry and

tectonic setting of magmatic units in the Pan-African Gariep Belt, Namibia.

Chemical Geology, 103, 101–121.

Fuller, C.W., Willett, S.D. and Brandon, M.T. 2006. Formation of forearc

basins and their influence on subduction zone earthquakes. Geology,

34, 65–68, doi: 10.1130/G21828.1.

Gaucher, C., Frimmel, H.E., Ferreira, V.P. and Poiré, D.G. 2004. Vendian-

Cambrian of western Gondwana: Introduction, Gondwana Research,

7, 657–660

Germs, G.J.B. and Gresse, P.G. 1991. The foreland basin of the Damara and

Gariep orogens in Namaqualand and southern Namibia: stratigraphic

correlations and basin dynamics. South African Journal of Geology,

94, 159–169.

Goscombe, B. and Gray, D.R. 2007. The Coastal Terrane of the Kaoko Belt,

Namibia: Outboard arc-terrane and tectonic significance. Precambrian

Research, 155, 139–158.

Goscombe, B., Hand, M. and Gray, D. 2003. Structure of the Kaoko Belt,

Namibia: progressive evolution of a classic transpressional orogen. Journal

of Structural Geology, 25, 1049–1081.

Goscombe, B. and Gray, D.R. 2008. Structure and strain variation at mid-

crustal levels in a transpressional orogen: A review of Kaoko Belt structure

and the character of West Gondwana amalgamation and dispersal.

Gondwana Research, 13, 45–85.

Gray, D.R., Foster, D.A., Meert, J.G., Goscombe, B.D., Armstrong, R., Trouw,

R.A.J. and Passchier, C.W. 2008. A Damara Orogen perspective on the

assembly of southwestern Gondwana. In: R.J. Pankhurst, R.A.J. Trouw,

B.B. de Brito Neves and M.J. De Wit (Editors). West Gondwana: pre-

Cenozoic correlations across the South Atlantic region. Geological Society,

London Special Publication, 294, 257–278.

Gresse, P.G. 1995. Transpression and transection in the late Pan-African

Vanrhynsdorp foreland thrust-fold belt. South Africa, Journal of African

Earth Sciences, 21, 91–105.

Gresse, P.G. and Scheepers, R. 1993. Neoproterozoic to Cambrian

(Namibian) rocks of South Africa: A geochronological and geotectonic

review. Journal of African Earth Sciences, 16, 375–393.






71

Hälbich, I.W. and Alchin, D.J. 1995. The Gariep belt: stratigraphic-structural

evidence for obliquely transformed grabens and back-folded thrust stacks

in a combined thick-skin thin-skin structural setting. Journal of African

Earth Sciences, 21, 9–33.

Hartnady, C.J., Newton, A.R. and Theron, J.N. 1974. Stratigraphy and

structure of the Malmesbury Group in the southwestern Cape. Bulletin of

the Precambrian Research Unit, University of Cape Town, South Africa, 15,

193–213.

Ingersoll, R.V. 1979. Evolution of the Late Cretaceous forearc basin, northern

and central California. Geological Society of America Bulletin,

90, 813–826.

Jordaan, L.J., Scheepers, R. and Barton, E. 1995. The geochemistry and

isotopic composition of the mafic and intermediate igneous components

of the Cape Granite Suite, South Africa. Journal of African Earth Sciences,

21, 59–70.

Kimura, G., Kitamura, Y., Hashimoto, Y., Yamaguchi, A., Shibata, T., Ujiie, K.

and Okamoto, S. 2007. Transition of accretionary wedge structures around

the up-dip limit of the seismogenic subduction zone. Earth and Planetary

Science Letters, 255, 471–484.

Kisters, A.F.M., Belcher, R.W., Scheepers, R., Rozendaal, A., Jordaan, L.S. and

Armstrong, R.A. 2002. Timing and kinematics of the Colenso Fault: The

early Paleozoic shift from collisional to extensional tectonics in the Pan-

African Saldania Belt, South Africa. South African Journal of Geology,

105, 257–270.

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.

Little, T.A., Cox, S., Vry, J.K. and Batt, G. 2005. Variations in exhumation level

and uplift rate along the oblique-slip Alpine fault, central Southern Alps,

New Zealand. Geological Society of America Bulletin, 117, 707–723.

Merle, O. 1997. Strains within thrust-wrench zones. Journal of Structural

Geology, 19, 1011–1014.

Miller, R.M., Frimmel, H.E. and Will, T.M. 2009. Geodynamic synthesis of the

Damara Orogen sensu lato. In: C. Gaucher, A.N. Sial, G.P. Halverson, H.E.

Frimmel (Editors). Neoproterozoic-Cambrian tectonics, global change and

evolution: a focus on southwestern Gondwana. Developments in

Precambrian Geology. Elsevier, the Netherlands, 16, 231–235.

Moore, G.F., Park, J.-O., Bangs, N.L., Gulick, S.P., Tobin, H.J., Nakamura, Y.,

Sato, S., Tsuji, T., Yoro, T., Tanaka, H., Uraki, S., Kido, Y., Sanada, Y.,

Kuramoto, S. and Taira, A., 2009. Structural and seismic stratigraphic

framework of the NanTroSEIZE Stage 1 transect. In: M. Kinoshita,

H. Tobin, J. Ashi, G. Kimura, S. Lallement, E. J. Screaton, D. Curewitz,

H. Masago, K.T. Moe and the Expedition 314/315/316 Scientists (Editors.),

Proceedings of the Integrated Ocean Drilling Program. Vol. 314/315/316.

Integrated Ocean Drilling Program Management, International, Inc.,

46, doi: 10.2204/iodp.proc.314315316.102.2009.

Moore, J.C. and Allwardt, A. 1980. Progressive deformation of a Tertiary

trench slope, Kodiak Islands, Alaska. Journal of Geophysical Research,

85 (B9), 4741–4756.

Nakashole, A.N. 2004. Sedimentology of the Malmesbury Group’s Tygerberg

Formation on Robben Island, off Cape Town. Unpublished BSc Honours

thesis, University of Cape Town, South Africa, 41pp.

Norton, I.O. and Sclater, J.G. 1979. A model for the evolution of the Indian

Ocean and the breakup of Gondwanaland. Journal of Geophysical

Research, 84 (B12), 6803–6830.

Oyhantçabal, P., Siegesmund, S., Wemmer, K., Presnyakov, S. and Layer, P.

2009. Geochronological constraints on the evolution of the southern Dom

Feliciano Belt (Uruguay). Journal of the Geological Society, London, 166,

1075-1084, doi: 10.1144/0016-76492008-122.

Park, J.-O., Tsuru, T., Kodaira, S., Cummins, P.R. and Kaneda, Y. 2002. Splay

fault branching along the Nankai subduction zone. Science,

297, 1157–1160.

Porada, H. 1979. The Damara-Ribeira Orogen of the Pan-African-Brasiliano

cycle in Namibia (Southwest Africa) and Brazil as interpreted in terms of

continental collision. Tectonophysics, 57, 23–265.

Porada, H. 1989. Pan-African rifting and orogenesis in southern to equatorial

Africa and eastern Brazil. Precambrian Research, 44, 103–136.

Rapela, C.W., Pankhurst, R.J., Casquet, C., Fanning, C. M., Baldo, E.G.,

González-Casado, J.M., Galindo, C. and Dahlquist, J. 2007. The Río de la

Plata craton and the assembly of SW Gondwana. Earth-Science Reviews,

83, 49–82.

Reid, D.L., Erlank, A.J. and Rex, D.C. 1991. Age and correlation of the False

Bay dolerite dyke swarm, south-western Cape, Cape Province. South

African Journal of Geology, 94, 155–158.

Rozendaal, A., Gresse, P.G., Scheepers, R. and Le Roux, J.P. 1999.

Neoproterozoic to early Cambrian crustal evolution of the Pan-African

Saldania Belt, South Africa. Precambrian Research, 97, 303–323.

Sample, J.C. and Reid, M.R. 2003. Sedimentogical evidence from the

Maastrictian Kodiak Formation (Southern Alaska) for large-scale uplift

along the northeast Pacific rim related to transpressional tectonics.

Abstracts with Programs – Geological Society of America Fall Meeting,

35 (6), 547.

Sanderson, D.J. and Marchini, W.R.D. 1984. Transpression. Journal of

Structural Geology, 6, 449–458.

Scheepers, R. 1995. Geology, geochemistry and petrogenesis of Late

Precambrian S-, I- and A-type granitoids in the Saldania belt, Western

Cape Province, South Africa. Journal of African Earth Sciences, 21, 35–58.

Scheepers, R. and Armstrong, R. 2002. New U-Pb SHRIMP zircon ages of the

Cape Granite Suite: implications for the magmatic evolution of the

Saldania Belt. South African Journal of Geology, 105, 241–256.

Scheepers, R. and Poujol, M. 2002. U-Pb zircon age of Cape Granite Suite

ignimbrites: characteristics of the last phases of the Saldanian magmatism.

South African Journal of Geology, 105, 241–256.

Seeber, L., Sorlien, C., Steckler, M. and Cormier, M.-H. 2010. Continental

transform basins: Why are they asymmetric? EOS, 91 (4), 29–30.

Smallshire, R.J. 1999. The kinematics of soft-linked thrust systems.

Unpublished Ph.D. thesis, University of Leeds, School of Earth Sciences,

U.K.

Smith, A.G. and Hallam, A. 1970. The fit of the southern continents. Nature,

225, 139–144.

Solari, L.A., Torres de Léon, R., Hernández Pineda, H., Solé, J., Solís-

Pichardo, G. and Hernández-Treviño, T. 2007. Tectonic significance of

Cretaceous–Tertiary magmatic and structural evolution of the northern

margin of the Xolapa Complex, Tierra Colorada area, southern

Mexico. Geological Society of America Bulletin 119, 1265–1279,

doi: 10.1130/B26023.1.

Stanistreet, I.G., Kukla, P.A. and Henry, G. 1991. Sedimentary basinal

responses to a late precambrian Wilson Cycle: the Damara Orogen and

Nama foreland, Namibia. Journal of African Earth Sciences, 13, 141–156.

Stevens, G., Villaros, A. and Moyen, J.-F. 2007. Selective peritectic garnet

entrainment as the origin of geochemical diversity in s-type granites.

Geology, 35, 9–12.

Tankard, A., Welsink, H., Aukes, P., Newton, R. and Stettler, E. 2009. Tectonic

evolution of the Cape and Karoo basins of South Africa. Marine and

Petroleum Geology, 26, 1379–1412.

Theron, J. N., Gresse, P. G., Siegfried, H. P., Rogers, J. 1992. The Geology of

the Cape Town Area: Map Explanation Sheet 3318 (1:250 000). Geological

Survey of South Africa (Counsel for Geoscience), South Africa. 140pp.

Tikoff, B. and Peterson, K. 1998. Physical experiments of transpressional

folding. Journal of Structural Geology, 20, 661–672.

Titus, S.J., Housen, B. and Tikoff, B. 2007. A kinematic model for the

Rinconada fault system in central California based on structural analysis of

en echelon folds and paleomagnetism. Journal of Structural Geology,

29, 961–982.

Tokuhashi, S. 1989. Two stages of submarine fan sedimentation in an ancient

forearc basin, central Japan. In: A. Taira and F. Masuda (Editors),

Sedimentary facies in the active plate margin. Terra Scientific Publishing

Company, Tokyo, Japan, 439–468.

Trop, J.M. 2008. Latest Cretaceous forearc basin development along an

accrettionary convergent margin: South-central Alaska. Geological Society

of America Bulletin, 120, 207–224.

Trop, J.M., Ridgway, K.D., Sweet, A.R. and Layer, P.W. 1999. Submarine fan

deposystems and tectonics of a Late Cretaceous forearc basin along an

accretionary convergent plate boundary, MacColl Ridge Formation,

Wrangell Mountains, Alaska. Canadian Journal of Earth Sciences,

36, 433–458.

Unruh, J.R., Dumitru, T.A. and Sawyer, T.L. 2007. Coupling of early Tertiary

extension in the Great Valley forearc basin with blueschist exhumation in

the underlying Franciscan accretionary wedge at Mount Diablo, California.

Geological Society of America Bulletin, 119, 1347–1367.

Valensise, G. and Ward, S.N. 1991. Long-term uplift of the Santa Cruz

coastline in response to repeated earthquakes along the San Andreas fault.

Bulletin of the Seismological Society of America, 81, 1694–1704.

Villaros, A., Stevens, G. and Buick, I.S. 2006. Origins of the S-type Cape

Granites (South Africa). Goldschmidt Conference Abstracts (A673),

doi: 10.1016/j.gca.2006.06.1259.

Von Veh, M.W. 1982. Aspects of the structure, tectonic evolution and

sedimentation of the Tygerberg Terrane, southwestern Cape Province.

Unpublished MSc thesis, University of Cape Town, South Africa, 97pp.

Wakabayashi, J., Hengesh, J.V., Sawyer, T.L., 2004. Four-dimensional

transform fault processes: progressive evolution of step-overs and bends.

Tectonophysics, 392, 279–301.

Editorial handling: J.M. Barton



structural geology of robben island : implications …crowe/downloads/rowe_etal_2010.pdf · south...

Documents