Basement control and transfer tectonics in the Recôncavo-Tucano-Jatobá rift, Northeast Brazil

Tectonophysics,

41

154 (1988) 41-70

Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

Basement control RecBncavo-TucanoEDISON ’ PETROBM’S,

and transfer tectonics in the JatobA rift, Northeast Brazil

JOSE MILAN1

Departamento

de Exploraqio,

65-130 andar-20.035-Rio ’ Curso de Pbs-Graduaqio

’ and IAN DAVISON

2

Setor de Bacias Interiores, Avenida Chile, de Janeiro, RJ (Brazil)

em Geologia, UFBA, Rua Caetano Moura, 123-40.210-Salvador,

BA (Brazil)

(Received August 24, 1987; revised version accepted January 25,1988)

Abstract Milani, E.J. and Davison, I., 1988. Basement control and transfer tectonics in the Recancavo-Tucano-Jatoba Northeast Brazil. Tectonophysics, 154: 41-70.

rift,

The Re&tcavo-Tucano-JatobP rift consists of a series of asymmetric grabens which are separated by basement highs and transfer faults. Opening of the rift took place in a NW direction oblique to the N-S rift trend. Well defined transfer faults parallel the opening direction. They were responsible both for offsetting en-echelon depocenters in the Tucano and Reci3ncavo basins and for switching of the asymmetry of half-grabens across the Vaza-Barris fault zone. The transfer faults show characteristic features such as change of movement sense along strike and with time, and “cactus-shaped” fault structures as well as flower structures in cross section. The sigmoidal plan-form of the rift is probably due to faulting following preexisting weaknesses through a complex mosaic of basement blocks. Basement anisotropy is thought to have controlled the asymmetry of the half-grabens and the localisation of the Vaza-Barris transfer. A 2O anticlockwise rigid rotation of the East Brazilian Microplate relative to the S%o Francisco Craton around a pole located near the eastern termination of the Jatoba Graben describes the calculated 20% extension in the South Tucano and Re&ncavo grabens, the oblique northwestward opening, and the eastward shallowing of the Jatoba Graben. Gravity modelling suggests important crustal thinning, locally up to 458, below the rift. Mantle upwarping is localized near the faulted margin of the rift. With localized thinning in a 100 km wide basin during a 20 Ma rifting event, lateral, as well as vertical heat conduction would be important and may account for the absence of a post-rift thermal subsidence phase.

Introduction

resulted in proven recoverable reserves of 250 million m3 of oil. Seismic reflection data are limited

The Re&rcavo-Tucano-Jatoba Cretaceous rift system is located in northeast Brazil (Fig. 1). It is an aborted intra-continental rift, filled with nonmarine sediments, that opened in Late Jurassic to Early Cretaceous times during South Atlantic rifting. Gil exploration has been intensive since the first discoveries in the early 1940’s. Over 1000 exploration wells have been drilled, which have

to the Rec&rcavo and South Tucano basins where known hydrocarbon reserves are concentrated (Fig. 1). In the rest of the Tucano and Jatoba basins only gravity, magnetics and seismic refraction data and information from 15 stratigraphic test wells are available. Ghignone and Andrade (1970) reviewed the structural evolution of the Reconcavo Basin and

004O-1951/88/$03.50

0 1988 Elsevier Science Publishers B.V.

42

\ \ \ i\

r OILANDGAS ACCUMULATJONS

~ Fig. 1. Location

map of the RecZmcavo-Tucano-JatobA

rift system with maJor hydrocarbon

occurrences

shown.

concluded that this part of the rift was subjected to two discrete extensional faulting phases: the

ment structure, which is believed to have played a decisive role in the basin’s architecture.

first, Berriasian phase, during the deposition of the Candeias Formation and a second, post-Barremian phase, after the deposition of the SZo

Rift stmtigraphy

Sebastigo Formation. Subsequently, structural models have become more elaborate and diverse (Netto 1978; Ghignone, 1979; Cohen, 1985; Ussami et al., 1986). However, all models are based on individual basins within the rift system, or have focussed on particular aspects of tectonic evolution, with limited access to data. This paper presents a tectonic model for the evolution of the whole rift system, using all the published and PETROBRbtS proprietary data. The model presented here is integrated with surrounding base-

There are localized outcrops of thin continental Carboniferous and Permian sediments on the rift margins and in the Santa BrIgida Graben (Fig. 1). Further study is required to explain their occurrence. Pre-rift Upper Jurassic to Lower Cretaceous continental red beds (Brotas Group, Itaparica Formation and Taua Member of the Candeias Formation) reach a thickness of 1.5 km in the southern Re&ncavo Graben and thin progressively north to the JatobA Graben. The original extent of the pre-rift sedimentation was much

43

larger than the present-day rift area as these sediments are found in the Gabon and Cabinda basins of Africa (Netto and Oliveira, 1985). No evidence of syn-depositional faulting has been observed in these sediments. The rift phase was implanted during the Valanginian when faulting occurred along the eastern margin of the Reconcavo and Tucano grabens, and great thicknesses of fanglomerates were deposited (Salvador Formation). A deep lake developed and shales of the Gomo and Maracangalha units (main hydrocarbon source rocks) were deposited along with occasional turbidite influxes and sandstone fan incursions (Netto and Oliveira, 1985). In the Hauterivian, the subsidence rate declined and sediments filled the lake with prograding delta fans (Ilhas Formation). The rift phase terminated with fluvial deposits of the Marfim and SZo Sebastigo formations. During the Aptian an unconformable post-rift phase of fluvial conglomerates (Marizal Formation) developed, and no further sedimentation is recorded after this. In the North and Central Tucano and Jatoba grabens, sandstones are much more frequent than in the Reconcavo Graben, indicating the sense in which the rift was filled. Correlations are not always easily made, but the general stratigraphic

E

Wu.Il

CHRONOSTRATIGRAPHY

1

LITHOSTRATIGRAPHY

INTERNAT. NOMENCL APTIAN

framework holds for all the rift system. A schematic stratigraphic column is presented in Fig. 2 (Netto and Oliveira, 1985). Rift architecture

The rift system consists of a series of four asymmetric grabens: (1) Reconcavo, (2) South and Central Tucano, (3) North Tucano and (4) Jatoba (Fig. 3). Rechcavo

Graben

The South Tucano graben is separated from the Reconcavo graben by the basement Apora High (Fig. 4), which continues south as the Boa Uni~o and Dom Jo20 Highs, where the basement is within 1 km of the surface. These highs also divide the Reconcavo into eastern and western sub-basins (Fig. 3). The western sub-basin is structured with a NW-dipping extensional fault system (Fig. 8). The fault planes dip around 60” and have planar domino style profiles to the depth of seismic resolution (crystalline basement). There is no obvious roll-over or differential rotation of bedding between fault blocks. The eastern sub-basin also has a SE-dipping basin floor but with SE-dipping

BASIN

MARIZAL

119

M_W”I~,L SEDIMENTS

FLUVIAL

SYSTEM

125

MARACANGALHA TURElDlTES

Ee

DA

GOMO TUREIDITES

SERRA

FLUVIAL, ALLUVIAL FLUVIAL LAKES,

Fig. 2. Stratigraphy nomenclature

of the Recancavo

and PETROBR&

Graben (modified

nomenclature

from Netto and Oliveira, 1985). The correlation

is approximate.

ITAPARICA FANS, PLANE,

SERG! ALlAyA

AFLIGIDOS

between the international

Fig. 3. Isometric diagram of rift architecture and basement structure.

extensional faults (Fig. 5). Close to the southeastem margin of this sub-basin, NW-dipping normal faults produced the ascetic depocenters of Camaqari, Miranga and Quiambina (Fig. 4). The maximum frequency of fault directions throughout the Reconcavo is N30°E. This fault set is the main extensional system with large stratigraphic growth components of the Candeias and Salvador formations identified across the faults (Fig. 5). Transfer faults in the east Reciincavo Graben

The other less dominant fault set trends N30” W (Fig. 4). All three faults (Mata-Catu, Itanagra-Arac;as and Pahneiras) show characteristics of strike-slip faults (Christie-Blick and Biddle, 1985; Harding, 1985). They are interpreted as transfer faults (Gibbs, 1984). The first two faults separate the eastern RecGncavo Graben into three distinct compartments, each one with an individual tectonic history.

The Mata-Catu transfer fault (M-CT) terminates to both the northwest and the southeast against N30”E trending normal faults, the throw of which changes drastically across the M-CT. At the northwest end, the boundary fault of the Dom Joao-Boa UniZ%oHigh shows a much larger throw northeast of its mutual intersection with the M-CT than to the southwest. This movement occurred principally during deposition of the Silo Sebastiao Formation, producing the Alagoinhas depocenter (Figs. 4 and 6). Such a throw change across the M-CT affecting the S;iio Sebastigo Formation indicates transfer of a dextral strike-slip component along the M-CT during deposition. At the southeastern end of the M-CT, the structural arrangement is different. Southwest of the M-CT, the Salvador Fault lies farther northwest than it does northeast of the transfer (Fig. 4). This indicates a sinistral strike-slip motion along M-CT in this region. This illustrates an important feature of transfer faults, in which the movement sense changes from

/ W -BASIN

BOUNDARY

-GEOLOGICAL

CROSS

SECTION

\ 747 )”

-TRANSPRESSIONAL -NORMAL TRANSFER

ZONE

FAULT FAULT

ANTICLINE OEPOCENTER

Fig. 4. Structural map of the ReCancavo and South Tucano grabens (PETROBRk$‘DEXBA, 1985). Rose diagrams show the cumulative length of faults in 5 o intervals that cross-cut the basement and the pre-rift sedimentary sequence, defined from seismic reflection data. Numbers refer to wells in Fig. 6.

B

Fig. 5. Schematic

geological

section

across

REC8NCAVO

the Carnape

GRABEN

depocenter of the Rechcavo

Graben. Location

B’

--I-

Fm

in Fig.

4.

Group

Well location

Elrotos

S. Amaro

_.-cl IINn Group

Salvador

47

offsets in developing depocenters where upper crustal extension is at a maximum. The Itanagra-Ara@ Fault shows a sinistral offset of the southeast border fault of 3 km, but no offset is shown against the northwest termination at the Boa UniHo High. The overall movement sense is difficult to determine. The shales of the Candeias Formation form diapirs in the Miranga depocenter but not in the Quiambina depocenter, on the other side of the transfer. The Palmeiras Fault is a minor feature but shows characteristics interpreted as indicating strike-slip motion in seismic profiles (Fig. 9).

original

dextral (in the northwest) to sin&al (in the southeast) along strike. The Movement sense may also change through time along the same segment of the fault, but this will be harder to recognize. In the central region of the M-CT, a zone of transpression was identified in seismic sections. Thrust repetition has been identified in wells, and surface folding mapped in the Sslo SebastiZo Formation. This is compatible with dextral strike-slip with transpression at a restraining bend (Fig. 4). The Camaqari and Miranga depocenter axes are each marked by a line of shale diapirs which show a dextral offset of approximately 15 km. This is an original depocenter offset because analyses of seismic profiles suggest a difference in extension of only about 3-5 km; a differential extension greater than 15 km would be required to produce the 15 km offset of depocenters because there is no offset of the northwest basin margin across the M-CT. This illustrates another characteristic of transfer faults, that they commonly accommodate

Olindina

- lnhombupc

South and Central Tucano grabens

The South and Central Tucano grabens are separated by the Itapicuru transfer fault but their tectonic styles are similar (Fig. 3). The basin floor dips consistently to the southeast with depocenters situated along the eastern border of the rift (Figs.

Alogoinhos

Mironga r

Jlqulo

lx_. m

Buroc~co

:zr

q

Arotu

144_

Vl u

&.’

-I- . ’

Fig. 6. Stratigraphy

R “PRE-P’FT” in the main depocenters

Rin ..”

Dom

J

“_

_~

Jooo

Ouiombino

Comagari I

6

7

?

/ ( Tatol depth L of well

+ rlrj fjcrm

5

-

s

Estimoted + basement Estimated Top Dom Jo&

of the South Tucano and Recbncavo

grabens. The location of wells is shown in Fig. 4.

48

D

SOUTH TUCANO GRABEN

Fig. 7. Schematic

geological

section

4 and 7). The main extensional N25”E,

dipping

across

the South Tucano

REC8NCAVO

and Rec6ncavo

fault system trends

to the northwest

(Fig. 4). Reflec-

tion profiles in the South Tucano Graben indicate a planar domino-style faulting, with no differential block rotation identified. The throw of the main southeast boundary fault zone (Inhambupe Fault, Fig. 4) of the South Tucano Graben increases northwards to Crisopolis where up to 8 km of sediments are seen on seismic profiles, and almost 5 km of conglomerates were encountered

grabens.

Legend:

western

North Tucano Gruhen

of the basin

shows

down-flexing

The

data are not available.

(Fig. 7).

The Itapicuru transfer fault

structures a special

in Fig. 4

North of the Itapicuru Fault, the depocenter has been mapped using gravity (see below), refraction data, and surface geology and lies along the eastern border of the rift (Fig. 13). The structure at depth is poorly constrained as most of the basin is covered by post-rift Marizal Formation and reflection

border

see Fig. 5. Location

D’

formed only when opposing dips of normal faults occur on either side of the transfer fault, halfcactus structures are probably more common (Fig. 12).

without penetrating the fan base (Fig. 10) in PETROBRbrS well l-BRN-l-BA (Fig. 6). The opposite with very little faulting

GRABEN

Itapicuru transfer fault exhibits flower and transpressional folding (Fig. 4) and type of structure for which we propose

the new name “cactus structure”, for obvious reasons (Fig. 11). Cactus structures differ from flowers in that the upper part of the branching faults are listric; unlike flowers, whose branching faults are generally convex with shallower angles near the surface. We interpret the cactus branches as normal listric faults, which have been cut by, or merge with, the transfer fault, whereas flower “petals” are generally oblique-slip with minor reverse or normal components. The lower cactus branches should have shallower dips than the upper branches if a common detachment is shared by the listric faults. A full cactus structure will be

The North Tucano Graben is separated from the Central Tucano by an important NW-SE trending structure known as the Vaza-Barris Arch (Fig. 13). To this date, no seismic reflection data have been shot across this arch, but semi-detailed gravimetric data are available, and wells and surface geological mapping define it as a structural arch with shallow basement. Despite the NW-SE trend of the arch, there are very few faults which are mapped over it with this trend. Almost all the surface faults trend approximately N30”E and are normal faults with down-dip slickensides. Surface geological mapping and gravimetric data indicate that the asymmetry of the rift flips across the Vaza-Barris Arch, with the basin depocenter on the east side of the Central Tucano Graben switching to the west side of the rift in the North Tucano Graben. Surface mapping of the

pp. 49-52

A'

Fig. 8. Interpreted seismic section showing the structural pattern of the Palmeiras fault zone. PETROBRAS line 26-RL-730. Location in Fig. 4.

Fig. 9. Interpreted seismic section across the western Rcconcavo Graben. PETRO BRAS line 26-RL-784. Location in Fig. '>.

pp. 53-54

Fig. 10. Interpreted seismic section showing the conglomerate fan that borders the steep dipping Inhambupe Fault. PETRO BRAs line 26-RL-1079. Location in Fig. 4.

I

V)

'J:i V) V)

ci.

0.

~ -'= ... ,\~

·Y·-Il':

-.-of;"'I\r':'.,

i

_

••

,=-~:

\1['-'-""~ ~".", -1"~"~-f'F''''.)'

}1l.

~.} II.~~ 'I¥~r>;-.,-.'I-' '/1. tl'\ft~; ~ t1 ;..!}j--.:~~ {~;;;-.~.... ~·~t~l'\&a',f'I-'~it·~· ~~l. ~ ~ ..'\:.fC~'~".'j-'=:'~:i ~.'I-'~' e:::t . . ~~ .:"""~-;;:<{ - - .... E r;'""'t-~'l :t . . -~~!:-;:;, 1'-: . ; . N ~~. "f~;}'.s1 f....\;&~~... .t.-h~"\\;-~t;.{((~~~:.~~ r~· __ ,_~~ '~ro~I!'\i~J..U~:';:\~:'i;;/c~\~\·\~i}~~f"\1\ _l,,!!~J,. .. ·d-_.s _\_~_.a ~ H.t,r~It._a~ .. a~r,:r ..

... . . -.-.. ~E#,1h.~; ):w/A\'in.!.d.~.·h~X:l~%-;loJld~"£iH~·~~~\--¥'nWf'i .::-.r~"'1\ ;-r-l\;-"''1.i;.~f- -::'-:f":.fl~- (;~;: "l'-\"~l.-.!'_ '"lt~·.-:=t-

~""i

• a,,·-[- .

tr-...

.....

,':::':../

\l'!(l,i1ot';

lt~ c , i\:t·'\'l,,';·"1\,;a •.• c.:.~!4 ";1·?t:'.--11·'.f:...-r~tt"..;

_.,.o?,\

~ ,d~·'f{}.M. ' t\'HL~~~'\';';\HN.wl-!lIjFff"!~f~'11~~llE-TJ":F~.~~~~" . ' ' : ' l i1' ~l\\P'. ;'''··\-;;'.1: h~''''iHli'-\B ::.~, 'q; ~r-"" J, ~l ~ •. td:-·:W~' '~.~·i1-i.~,:-\Vf ~·t"t~·~·\f~i"'-\:D~;{!:

4'"

(

·l~' ~:!lj·Ef'

·~-~~";.{1~ ~

1~""','l·f-,;4)..

=.-""",,,

il

'l.

t~ ~- .c:.A'.-:~~

[ =:= ,*..:.r

I~--"'-" ~-illl~~'

f];w.:I

"':".·b\l;l.'l1i\

Ii,\l

l.r\~.J~\·hi

. '. ;.t:~·~l' -'--j)- -:.- ,-,1 ~ ~F::" ~ ~J~ .~, 1 . ?~T"" =i':_~' f~~~a"i"~-~ ,cJ'J..' . :",;-0 iii;L 'f ·'11')\1 -J!.!' .• "-'x, ~. 'i}:~i'i""-';""J -;';"'\1 'l -,. ·-:~~·i*~}~E"f'-': '\\Wrt,yt~~?,~Mhl~;~".rt.tf\a~.·.:~~';i~~~f~~9~~~ .. t '- .,. ,." -- - r . . 1.t .{ . ",..... t: -.... ..... t;. ~ _ . } - - • -t P·:~'l~ ~:.,i~-;_~=j~~~ :_~·~;-PJ·~.pi~:~{f4·~··_~':;~~~~i_:.tf

c

f·~t%l::~ . ,~~ (iY~\1Jt~lf1W:·ftH\tll1i~l.~~}'":{~V;rf~.~h\ I, c~,,~, ~" I c~1!;'11_;'~l\~\~'iJh\\lil'-q;-~-I!i;-btJ;d.vt~.~ rtl 11-'-r\! it 1i~:~{~t, : 'J,c ~'mrt1fJ}hjj:~\~!~t~l~f'~ \It"J!rlf}:~:~tlJ ' " ' ~ ' ; 1 !.,~' ft· LI.lL:;llfll-".l.I''''.,I< \ .'." ,,1 . J.~ ~:f.-.,.. ,;.~' ft ~'I.r~.ii)~:m;i~",~~-l ~.'~~iil-'~'f'kJ'i1.it.-JJm-·'i-L.~I'~ .(;iln,;,l~ :Jd~.'ljY-i ~ ,Jtt....'. '\' k'i"I~;;:.~

\

~l'~/?J-'~11' ~..... .- • -- i,' - i . . . =-:~!;J\'"-i;"'I!4

;:

0

__" _ '

.\

~--c_-~~~~~~t~'l"i

f.'!·:''',,,,,~,,·o- 'f" :--~t" y~.n

.~

r.. . A.{. -"f irJ:£"4."5. l'!-li!f ,'\'\\H,- I';:"", :n,-it' E-·."'C;"t "'1-:. , .•. ~ '_0', i"':!"l"~'~"" ..~,.\", K\'J~11"','I~'I'" ~ .t~~l_'.:l:':'; ,'.J,;:.\:\H.-:,( c!l-j ~ "~l\~\C "~~1"';':-':'1.:\'\) ~VfHt) ~- _~:~~ = co Jc;1 . -'iO'l' .1. \~\l·"'·\l'c':;:""'." -\! \ '- ,'. 'i·n ~."i". ·,!;,y}.:;:~~J.-·'

.f~. ~,z:-_~~{·-li~ .~.l; ~·.l. Il

:

:::..! Lh1li ...,. tj .. C_-:.io>· ~.r. .: , " ••' ~'.:f.ti\~. }\ ,~,n._tj ~.~}.\ . ,-:~." '-.00'1(."If >I..C~~~'~1.,-:r.~4 . . "\..nf";If.~f

,-:

.~

-

-'-:

:,;r

-

'l;t

.r

':II.

r'l II.

lV,:,;"C.

:..-:·....-1.;:.:...

~li"'~~ .. ~t

-01"1-

i ,: , .. "- ... -} - \ . " -t·")

:"-

1:

z

.

t .•.

• . _,i-..:"'-

l.~

,,".

...

hi. =--h

a

.• a

...

:a_."T

.. _

" 'J.:.!

~-,,,,~,,,!<_.I!

k" .. &

...

I

li ~ ~;' [:.... ~,,'f inv 't", ~ i\Wf.\·c,,~;h", {~~,,'fi;l-' ~'l'l\' .' '.,;i~""";"-:' .•--""''fC~"'f-!: ·- -f' ......11"~l .., r:- /\-;,.."i·.... "~~·h'~,~~ll ~... ."·2\..,,..1 ~~ a:.1 &"2'~~ -~T." .. ·'1'11 f....... J.;..~~~ ....~: if/rot .~ .. ;.. ,.:::'-"'P .•. ~,,..,,,~h. ,v u.~ . . ,.~ •.-o..~":~,,.-;.~\:-:ta _-- -::..3:'.' ·"··1· .... -.·,1. .. :J.~,. ·r·' n '.!.. ~t.~ .. -.,_" .~ ~ .... .:;l,I-~ tl~.~.i~. ·J:I -~-=t~.~.t1·:i\I;~;'''!·i ;t.:.J 'til· .,.:-.-t----=j-~·.}:{i.ft. ~'"i-:-;;f ~-....... l!!i;\~£:=! .«t.<'1~f-;.-.~~~ f-'j ._~ ,'111·~·~t~:-i. ~Ji.ti.'~ -:.E':' -·,f ..,,- .-.£ _. I . ".:.iJ 'I J ~.:,~~.5J:nrfllij 1r-;:I~ll·~1~i~"2.hlh:·J.};~ ::;jt!r~p.c ..),.,1~"},iIi~·io.·~.·~Rf) 1. \:1~it~?,fNi _. H-,": I h\Yj~Yhjt'Nt!!{:"lr{d\~! HVi!tfi~; ~~!f~~{~ f .E. ¥~~j ~ ~; ;., J •. " II "[I' l\.:"''1Q ,'''if; ,.. ifl' ~ "VI .I",C"· rtf/~'!\-

.f} !~ %'} y~~ (M·tf.II~ ~H\! rm .~~l\~,tSf JU; \5 -f1 ~',~L ir"ien:tJ~~{;!f~r ~. ~'1'J-0' ' :'.I":.<'I .!~ti 't,:'-.'S~:;'~J'.~ !'fi>.I. 'f}I·.~ 4c.'!l r.l :I-"~h". - . - - - . - .-• .

';':."_ W

M ..

if'.:.l

f·f!~·r~

f/ef ltilii

a

-{-'~·:~-f~ll ~:~: ~lIJ~~! :t~:fl~~:, _{~J

tiltlt".f!~.'it~11i".:}'.~~ ~~.'~'#,f.f.§ l{fil~IL11:? 1;'~l'~I:"~"f'.:'~

··''-{b~ f f:.1l.,:E {-;'-1- :---,".! '1l~;,-T,H} '"1 / ; .~''';i.~U·"';..5:-1 fi ......!·\'l·\.'Y~ l<'. "':;i.' • 1~!"Ji'I""'1 -.,..

!

·L ~ tC., . ) I· . '1"·1 .~ Ii' ...-.-",. l '1"-'1-" -..,.. ..liJr"., l J.I\.~J_.1 t'L}'JLi. =-14.;, ~~ . ~-=~:5' "- .~_. L - . f-'"t ~;:~~:}J~~:'\-:~l~;;~t~ .~'(i';.. ;;:'~ ;;l:~if·lt ;,- ff-~~r;-~ .~-l.-~-~t~ i f.1-f • - ~". J'I.!f'kf~~fi~~'¥.in! f#P: {J; '~!1-f-f.'fljil,f.~.H.... ",;-f1F ~j '1"";"-'I~-,.tf'lf·li1fN" .t~·I·'~~10;f'1!-1~:1"1~-if::--ii:i~ '1-

~"i ~ ~ti1T'

i
131\.. ~

..

:J ,,;an-.f-(~ f:da"-~ 2·tl l - I 'if~;~f-~ll~!..?li~ :: ':.,. "f!{~Jl i 1: i~!- tJl1;:_~f.l~ 'lf~~ • 'W'. Zl-'~.". r·.·.· ··.t~ ~.t'l,rl'l'·_-:""'···'l"! .../',:.". ~l"~ \f~ t:-Hht.{!V_ U '.". b .. 'i?_~- \ .• : -;..f4:. ~U_i1,· - -t~:Ii.f ':"~.:t.!I~\\::':l ?;~~.l \:. ~3: ~.':{ t1a.. .~ "!., ti' :-':l·'--:"~· IiLll.,"~~---';''''''-{'':''L-ll·t1't:-t~'l~~'~' ~ "-l.. '-~t '", .--;:~ . . . -_~\""\i.~., . ~'r·c'll,·_·.:t!t;'5.·"",.-".d I,~ i{-·t~ ...."."~' .= ,-;;;... "L -. \.l\~'''l "~'7"'11oEJ'~"" "lll ·'i ~~ ~-"'O"'-''''''f.~-'''';'> p(tdrtf ~mtl~jl · t ~~~;"~U~'":~ .-~ ~- -'. ~'ijf:;"~l1 \~F ~ ~r.I,o,,,," ..-,.n:,""!»"'n-lU.) ".~"~\-f!-{~ '\:'1}.t~f1!f'·"!~ni~'h'1!Hflt~!. r~~I\f;t.~~,,~

-.' '··-=l~\·~Ill' <; .;;-.1

-.!

'!1

.-.-\=-.. -.' ."'.'.

<-~:".-~.~~ • ~ 2' ......... "' ~ .....,\.

-.-=

. § . ~{{,... ~ -

-s:i .~-~. \.> .. ..

lo::. hL

.S: - -

-

1l.... --.

i Y-i{!=-r'

-1tf 7i'

:i-·;"'·~fO::>~J,· ~"'~l""~~'" .:t £ - - _ ..1 J £!t . .. ... _.. H·",·,·'tt'i1E1-.l'. t""" ~"l ~

h\1

~_' l'~·o """'f·l.ffl-.\.·~{I;"'11IW-1.i.--4 ttt.. f'"~i':?i·:t~~~:;:, '~'~'\\:i-} ~~::.t4·r-l~}:~j',,'':I!il~n"l~h:h'~~J<'{1~~ f, . ' -~ '{~-'--':~-'1 ;""~ J ~ Ll ii!il~.t;';II.~.=.J··fl·H I

n

:'-~;J1 f~-·lf-.~ ;m'~'(\'1r "'-nHfJ 1,JH:1\11,. ~ir\\)l~}: -~.-:~.F -.-- ="';;:rJIIV~~lllf f}{;,l 1-;:i:f' Ull'ilff{ri~o~CJq~~\b-~f~-{·0fX~ ~---'~1t!'1 ~ .!-~ f·f·" Ult""'~-h:'-"ll!~l'{~i';)U='

.«

t

"ltt,N , .,,;:1 ,11'\\..1

{'Y,I"

'2

,lh..""

-

&

-

Y

•

-

•

:riC'

-

1

•

;

,-

-

"if-'l:-

t

a

•

..

.

!.

i:! ..

!ffi~A'i~-

ll-~l·Y~;' t WJJ,~I·~n>·~ 't~~.':,~!~ .~" !l~I:JJ'.~<'~_ih. J!:~JU''-i~f··'''-·-~.!''. f!h".,'; ~!iTh "H'£~'?J'.

~';'~Id,-~ !t~['illfJn.'~'t}jf,\:\1"'t/J';;

-~'1,10·.. -lr'-·3~··~J!}d,-l';."1F-~-'"

~,'t:.~,. . .T

?~,~I',c:l: ¥1~ t{

....

bl)

~ .~[,l.,

OS

l::

.S

_x ti+ i :-:, .... 1 Jt,!l~j.!.~:~fNf£~ Ir~.~ .. 3,Jfl.;~1l1E'}.rf~'k1'~J;'1tid'-f.19.!!;l::I~. .g !tJ'''''.·tIi!! g

1. . ':>JI' '.f 1f /1· !t:;I(:iYI.i·;j:";'iINJ{!;~·I~~c f If•.. ~ "'---"4~''''I' ,·.;::n I:_"~"'I~ ..•. .tt!{-'.~\t~ -~.. j ~f/t jJgtJI~W-A'!~5 -""n 'l};;!

- .i,..

..,,--;;-.:, J:'1·.-.. ~·-~·'l'r

-

_~f1 E.!\L~ . ,. J'IJ!-OOl."\! ,.,J.,1.,,~ ~f'"" • ;..;;,,: "-tiltfiw,."t011f;~!1"lillfL;~;~fh "-."0' ,--,"

~~: ~ ~;;~\~~1 W(i~.~~~·!f\r~1~ljl~~;'~1~~~ ~~

r-ll.

--

,r-"·~,'.'

--, 1.• ~ ':.1 .. , ,1- l'... ~ ~ "I -... -~, ·'i'-'" ·:-.r"·'-I4).-\.. L~""f,-,\l -f"j '-';'I};j, '1 f ..:J;l!··· ...i ;roi, --~ 1:,,·-.,· ~1r.~~I~'""., '-.= ~ ~f ::'~l·.~:~T=--;i l;:'~·. -<'f,pl~l: fi '-'I OlE '1'-'0,11;\

. . 1,I'''''';·1 .

-" == i. ==

-

.-. fE' }.!.,

~':r'C'\!~~~ ~i ~-.~ p-. 'i,;'.' it - "·1.~;'

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

:f-= ._""' 1 . F~ fi "'~

'"

1i:.

f~.

~

~ ~..

~?'l

i

O

t '

'E"';:i

'i

,:'-. .. !...

n:'r':E

"j"~;,u.~~;li:I;

".f.l-... j->t-,.;".--

i{·'":/"';!' ~r-'H"F,c,

L'tlMNa' .. J

~fJ~.;f~{rf/x·;~Yi

,,-i{

M

'
~~-~"3]I}JIJJ!Jhl C\!f fl}~i:d~lfl ;liJ'fjJJ U1}lj.~t~~I~~~Ji ~ lb;(qill-ct \l;1rAlp JHi{t/h:iF1 )l 3 ~"-rlIJ:w.rMi!~ 1i1)j"l[.k!i!/!i:Bj{fl{b~·l~~! s-.l-c -, -" .~l" f,!;t'(UffiUf3i!i, .~. Id!il1.F-c·. ~

l~ K~ ~1111~:1~iL ~fll1f-/llMtf~!n[tt:~:;}Wb1jHf~r/l~b:~tr} ]

O¥{

n

'::J:

_.

,.,.

•

11>

I-l"!

i. __

"1

••

!""_"1t

L'.J"".'

f/'IJ

i If~. i., -;J-..:~~- t.i";j_v-rlt;l~ Hf.!JlJA H.-.r.'.} f=.f;fF.,+~.nf~J..f~t;c·~.h{t- r/l :~ i-~~'J ::}: ~.; fl"i-:'ij-!1:!f4~JtJ :F'lJIJ,I'i".ii'W}/t}·J'~JI'WTt3 .
2:l

~I_ ~E-~~,,-~~.=:2·fl. ~1l'f}I[. 'A,~h__~~:!O_ Ii~'~g '£1'~Urr.~j~fI«"it1:rn!; : ~1:U,; <_i .7=~= ~

. l,.n

()

.s. ~

~

.s l::

o

'-

~

.g ()

'"

.~

'"

·S ~

]

~ oil

....i:: .... ....

LL:

51

fined by refraction data. The E-W trend of the basin is controlled by the Ibimirim Fault (Fig. 13) with the depocenter close to it. The depocenter shallows to the northeast where its northern border fault deflects 20° and merges with a major basement shear zone (Fig. 3). The southern and eastem borders of the Jatoba Graben are defined by gentle flexure without major normal faulting.

a

Gravity data

0

b

0

CACTUS

Fig. 12. Schematic map (a) and block diagram (b) of a cactus structure.

stratigraphic dips also shows this switch. Although there is no evidence of a simple fault structure trending NW-SE along the arch, we interpret the Vaza-Barris Arch as a transfer zone, similar to the accommodation zone defined by Bosworth (1985). Seismic refraction data from the North Tucano Graben indicate a more complex fold and fault geometry than in the other sub-basins of the rift (Fig. 13). Fold axes are variable and three fault sets are prominent: N15 ’ E, N85 o E and N25 o W. The N85”E fault set probably reflects controlling basement anisotropy. The variable fold axes indicate a complex compressional phase oblique to the margins. The northern border of the North Tucano Graben is bounded by the major N80 o E Ibimirim Fault (Fig. 13), which has a large downthrow to the south; fanglomerates were deposited along it. Gravimetric data also indicate a large depocenter along this border. Jatobci Graben

The Jatoba Graben is separated from the North Tucano by the Sgo Francisco Arch which is de-

The most distinctive features of the Bouguer anomaly map are the strong negative anomalies which run along the central part of the rift and the strong gradients along the major boundary faults (Fig. 14). The switch of the rift asymmetry across the Vaza-Barris Arch is very clearly defined. Using estimated densities of 2450 kg/m3 for basin sediments, 2580 kg/m3 for Palaeozoic sediments in the Jatoba Basin, 2620 kg/m3 for conglomerates of Salvador Formation, 2850 kg/m3 for the crystalline basement and 3300 kg/m3 for the mantle, closely constrained models of the upper crustal structure have been constructed (Fig. 15). The modelling method used a computer program derived from Talwani et al. (1959) technique which calculates the vertical and horizontal gravity components produced by 2-D polygonal models. Starting from a simplified 2-D geological model based on reflection and refraction seismic data and surface geological mapping, the program computes a synthetic gravity profile which is compared with the real profile. The model was adjusted iteratively, within the confines of the existing data, until the closest fit was obtained. The basin structure is well defined in profiles A, B, C, and D and therefore the main variable in the models is the positioning of the Moho. Four models with calculated best fit and real gravity profiles are shown in Fig. 15. Profile AA’ (Fig. 15a) shows the Jatoba Graben with a section perpendicular to the main gravity anomaly trend. The strong asymmetry of the basin with a northern dip of the basement is the main feature OTthis model. Profile BB’ (Fig. 15b) exhibits a similar asymmetry across the Central Tucano Graben. This NW-SE profile traverses the Sitio do Quint0 depocenter (Fig. 13), which is the deepest part of the whole rift system.

58

9%

90s

3RTH TU CAN0

\’

7

CENYRALTUCANO

-1,os

IMAPPED 8Y

Fig. 13. Structural the basement

map of the Central

level in 5 o intervals,

and North

defined

Tucano

and Jatoba

from seismic refraction

The 150 mGa1 negative anomaly suggests the basement is deeper than 12 km in this area. Profile CC’ (Fig! 1%) shows a NW-SE section across

South

Fig. 14. Bouguer

Tucano

anomaly

Graben,

and

profile

DD’

map of the RecZmcavo-Tucano-Jatobh

grabens.

Rose diagrams

data (modified

SEISMIC

REFRACTION1

show the cumulative

from Correia,

length of faults at

1965a. b).

(Fig. 15d) lies across the Camaqari depocenter of the RecGncavo Graben. The striking common feature of the four profiles is the offset

of the largest

rift system (PETROBRkS/RPBA,

1970)

negative

anomaly

59

390w

-- I

--

!8aW I

-J\,

37OW -j--B0

s

10

z

Oar

10 (mGal)

(Km)

A’

,

(Km) -45 4

40 RESIDUAL l------

1

(Km)_45

I

-

’

”

’

(Km) -45 1

”

”

”

”

d) 5 CmGal) (Km1 0’

Fig. 15. Two-dimensional observed

major border

faults. Location

a. Profile AA’, Jatoba Profile

gravimetric

profile. A corresponding

Solid lines are observed from Milani,

of profiles

Graben.

DD’, Rec&rcavo 1985, 1987).

models

geological b. Profile

across

crustal

the rift, with comparison

of the calculated

section is shown below each gravity

profile.

Rouguer Residual

gravity

profile

with the

curves show the position

of

and RecBncavo

d.

is shown in Fig. 14. BP’,

Central

Tucano

Graben.

c. Profile

CC’, South Tucano

grabens.

Graben. gravity

profiles

and dot-dashed

limes are calculated

profiles

derived

from the geological

models

(modified

61

from the deepest part of the sedimentary basin. Residual curves show the real position of major border faults that control the position of depocenters. There is also a consistent difference between the plateau anomalies on the opposing rift shoulders. These features require mantle relief similar to that shown in Fig. 15. The four profiles presented show that the calculated anomalies produced by the elevated Moho (crustal thinning) coincide with the area of deepest subsidence in the basin. This suggests that the isostatic compensation of the graben was a response to local mantle uplift. Ussami et al. (1986) proposed an easterly-dipping crustal detachment fault beneath the Central Tucano. They used a simplified Bouguer anomaly map, which does not show positive anomalies flanking the basin, and suggested that the absence of flanking anomalies implies that there was no important crustal thinning below the rift. However, our more detailed Bouguer anomaly map shows positive flanking anomalies up to 50 mGa1, which implies that crustal thinning along the northern and eastern margin of the rift was important. An easterly dipping detachment fault would be in direct conflict with the asymmetry of the Central Tucano Graben, where the major normal fault controlling basin asymmetry dips to the west. Crustal stretching

Crustal profiles for the South Tucano and Recbncavo grabens, similar to those shown in Fig. 15, permitted construction of a simplified depth contour map of the Moho, which indicates localised thinning anomalies next to the Salvador and Inhambupe faults (Fig. 16). Balancing of crustal cross sections perpendicular to the crustal thinning contours and the principal normal fault trend (i.e. parallel to the principal extension direction) should provide a rough quantitative estimate of the average crustal extension below the rift (Gibbs, 1983). Using the area balance principle on two sections CC’ and EE’, which traverse the South Tucano and Rec&cavo grabens in a NW-SE direction, and assuming that the original crustal thickness is equal to the pres-

ent-day crustal thickness on the rift shoulders, the average crustal extension was estimated at 20% (Fig. 17). Locally, the extension reaches 45% below the eastern margin of the rift. With the Moho anomalies similar to the ones modelled along profile CC’, one would expect significant lateral, as well as vertical conduction of heat into the rift margins during the rifting event. Although dating is not very accurate the ostracod biostratigraphy suggests that rifting occurred over a period of 20 Ma (Fig. 2). With such prolonged rifting, and a crustal thinning profile such as that shown in profile CC’ (Fig. 17) it can be shown that the heat anomaly created during rifting would be dissipated during the rifting event itself, so that no post-rift thermal subsidence would be expected (Co&ran, 1983). Post-rift sediments are absent indeed. Ussami et al. (1986) suggested that the absence of post-rift thermal subsidence in the Central Tucano Graben is another indication of a discrepant zone (Wernicke, 1985) and that rifting was affected by downslope sliding on a low-angle easterly-dipping detachment which links with the basal detachment surface below the Atlantic Ocean Basin. Opening direction

It was demonstrated that the principal extensional faults strike N30” E throughout the Re&ncavo, South and Central Tucano grabens and that the gravity data indicate that the depocenters within these grabens also have an approximate N30”E alignment. However, the general trend of the rift is N-S, suggesting that oblique rifting occurred. Withjack and Jamison (1986) showed experimentally that extensional faults do not always develop perpendicular to the extension but may vary up to 20” from this direction during oblique rifting. Thus, the major normal fault trend may be slightly. oblique to the opening direction in the rift system. The important N30°-40’ W transfer faults have steeply dipping to vertical profiles and show both extensional and compressional structures along their strike suggesting they lie close to the opening direction. One can therefore conclude that the rift opening direction probably lies be-

/i

D ./

, ,

1 \ \ ~

\_

Fig. 16. Contour map of the modelled depth to the Moho below the Recancavo and South Tucano grabens using four model crustal profiles (from Milani, 1987).

tween NM040 o W. This interpretation differs from Cohen (1985) who suggests that the opening of the RecanCavo Graben was oriented E-W. Cohen based his conclusion on the orientation of faults mapped at the surface in the Re&ncavo Graben and concltied that the N30° E faults

were strike-slip faults with large normal components of throw. However, we fail to see how the NE-~~~~g tr~~~~t shear zone, proposed by Cohen (1985), explains major NE-trending extensional normal faults which were active during sedimentation. Cohen also suggests that the

63

CRUSTAL STRETCHING 1, = 149

I

SOUTH

TUCANO

Km

AKd

HIGH

0

-4 0 Km

2yKm

, AI:27Km, 1, ~128

1 I

E

(e = 23%)

Km I1 ~184

I

Al=35Km

SOUTH

0

(e

= 21%)

Km

1

II =155Km TUCANO

REC8NCAvO

I

Km Km

lo =Original

length

I, = Post-rift

length

bl=lncremenf e = extension

af crustal

in black

black

length

I II_ 10

Fig. 17. Area balancing of two crustal sections, CC’ and EE’ (location on Fig. 16). The dotted area indicates the sedimentary basin fill and the diagonal lined area represents the original area of crust which was extended. Balancing assumed that the crust had an initial thickness equal to that on the undeformed rift shoulders (from Milani, 1987).

sigmoidal shape of the rift system indicates a northeast sinistral shear zone, but if the sigmoidal shape of the rift were to be explained as a mega tension gash, this would require a west-northwest sinistral or north-northeast dextral shear zone (Ramsay, 1980). The Reconcavo and Jatoba grabens should also be younger than the Tucano Graben, but this is not the case. An interpretation of the sigmoidal rift shape is presented in the following section.

Basement control The rift cuts a complex path through the mosaic of basement blocks which played a decisive con-

trol in the rift geometry. Data of penetrative schistosities, shear zones and brittle faults within the basement have been collected from existing geological maps and field work to evaluate their control on basin opening (Figs. 3 and 18). The eastern margin of the Recbncavo Graben is bordered by Early Proterozoic gram&e gneisses with a general foliation strike of N30 o E and steep NW dip of the gneissic banding. The Apora High is composed of Early Proterozoic amphibolite gneisses which are transected by N30“E, 80 “-90” NW oriented shears up to 1 km wide. The western border of the Rec6ncavo Graben is composed of granulite gneisses, but with a N30 o W

J

f-i

FOLIATION

TREND

BASIN SEDIMENTS ( J - K) DENSER DOTS = DEPOCENTERS TRANSFER AND STRIKE-SLIP FAULTS MOVEMEW : P - PREC&MERiAN K - EARLY CRETACEOL INFERRED DATA)

FAULTS

(GRAVIMETRIC

LATEPROTEROZOIC FAULTED

,..’

_

8;) / 25

50Km

/

worked (Late Proterozoic) gneisses and granites; 7-Lower Itapicuru Greenstone Belt.

Proterozoic granite-greenstone

FLEXED

RIFT

GNEISS

BANDING-

NON-

LP

_

LATE

EP

-

EARLY

Fig. 18. Structural map of the basement surrounding the rift (modified from Brasil/DNPM, 1981, 4 and 5-Late 2-Lower Proterozoic gram&es-Atlantic Belt; 3-Upper Proterozoic sediments;

RIFT

SPECIFIED

THRUST

MARGIN

MARGIN

SCHISTOSITY

FAULT

PROTEROZOIC PROTEROZOIC

and field work by authors). I and Proterozoic fold belts; 6-Re-

terrain; black areas correspond

to

the Rio

65

foliation trend, dipping 70”~80 ‘SW (Fig. 18). Farther north, between the towns of Serrinha and Euclides da Ctmha, Lower Proterozoic granitegreenstone terranes strike N-S with a 60” W dip. Along the eastern border of the South and Central Tucano Graben the basement is covered by a veneer of flat-lying Upper Proterozoic sediments which may reach up to 2 km thick south of the Itaporanga Fault. The underlying crystalline basement may also be an Lower Proterozoic granite-greenstone terrane with a similar N-S orientation. The Itaporanga, SHo Miguel do Aleixo and Belo Monte faults are all interpreted as sinistral transcurrent shear zones formed before and during the Brazilian (Late Proterozoic) orogeny which produced the Sergipano Fold Belt (field work by authors). All these faults are steeply N-dipping or vertical, and converge towards the Vaza-Barris Arch. They are probably deep-reaching crustal scale faults which separate stratigraphic terranes. The S5o Miguel do Aleixo Fault can be traced across the Tucano Basin and emerges near Canudos town (Fig. 18). The Itaporanga Fault may control the orientation of a transverse residual gravity anomaly which is interpreted as a strike-slip fault. The southeast continuation of the Itaporanga Fault is also responsible for the southwest termination of the Sergipe-Alagoas Basin (Fig. 18). The Sergipano Fold Belt consists of steep NNE-dipping folds south of the SZo Miguel do Aleixo Fault, and moderately NNE-dipping folds north of this fault. The Pemambuco-Alagoas Massif is dissected by a series of NE-SW trending shear zones and faults. The Ibimirim Fault parallels one such shear zone, but the most important shear zone is the Pernambuco Lineament which is the longest in Northeast Brazil. Drag structures and observation of stretching lineations indicate a transcurrent dextral shear movement of many kilometres, but towards its eastern end steep reverse faulting occurs. The age of the ductile transcurrent movement is thought to be Late Proterozoic, but the age of the localized brittle reverse faulting is probably Mesozoic, the same as the rifting episode. From the above disposition of basement planar weaknesses, it is clear that some of these have

been faithfully followed by the rift faults but other rifts cross-cut all visible basement structures. The rift system is an excellent example of nature’s way of finding the easiest path through the labyrinth of a complex basement structure. The faults which developed parallel to pre-existing basement weaknesses (schistosity, lithological contacts and major shears) are shown in Fig. 19. The sigmoidal shape of the rift shows a convincing fit to the surrounding basement trends (Fig. 18). The asymmetry of the rift is also thought to be due to basement anisotropy. The eastern depocenters in the Reconcavo, South and Central Tucano are caused by westerly dipping master faults parallel to the shear zones, gneissic banding and lithological con-

e--

FLEXED

-

MAJOR FAULTS TO BASEMENT

7

MAJOR FAULTS, NO BASEMENT CONTROL

0

sub-parallel

showing

to schistosity,

and faults within the basement.

PARALLEL ANISOTROPY

X)Km

Fig. 19. Map of the rift system developed

MARGINS

major faults that

gneissosity,

shear zones

66

tacts in the adjacent unfaulted

western

orientation orientation readily

border of the rift, whereas rift margin

unfavourable of the North

explained

exhibits

for reactivation. Tucano regardless

The N-S

Graben

in terms of basement

as the rift cuts through

the

a basement is not

weakness

of the NW-SE

oriented basement. However, the choice of the Vaza-Barris as the transfer zone to accommodate the switch of basin and Central basement Itaporanga.

Tucano faults The

asymmetry

between

grabens

was controlled

of %o abrupt

the North

Miguel

do Aleixo

and

northern

termination

of

Arcsctntrtd

on

Rtlotivt movtmtnf

171

Eattnsion oxit

,/’ 1//1

Tronsftr oxis

Thrust fault

12%

Inferred segment of fht •l2

Fig. 20. The East Brazilian microplate (hatched area) with rotation pole marked

(from Milk

by the

1987).

microplott boundary

the rift system against the Pernambuco Lineament is clearly basement controlled. Ussami et al. (1986) suggested that an easterly-dipping low-angle detachment may have been activated along the basal thrust plane of the Sergipano Fold Belt. However, a postulated N-S trending low-angle basal thrust bears little relation to the known geology of this fold belt (Figs. 3 and 18). All presently outcropping major thrusts and strike-slip faults dip steeply NE-NNE. One can easily conclude that no clear evidence exists to support low angle crustal detachment below the Tucano Graben. East Brazilian microplate

Szatmari et al (1985) suggested that a triangular-shaped crustal plate limited to the north by the Pernambuco Lineament, to the west by the opening axis of the Rec&cavo-Tucano rift system, and to the southeast by the Atlantic Ocean, existed during the Early Cretaceous (Fig. 20). They explained rift opening by horizontal body rotation of a rigid plate. This is rather a simplistic model as internal deformation of the plate also probably occurred, but it is difficult to separate intraplate Precambrian deformation from Early Cretaceous deformation. Nevertheless, the model fits closely to the observed rift geometry. Iterative positioning of the rotation pole describing the plate rotation gave a best fit position of 8”ll’S and 36 o 04’W (Fig. 20). The approximate 20% extension of the Reconcavo and South Tucano grabens (measured from section balancing) requires a 2” anticlockwise rotation of the rigid plate. A rigid rotation around the pole implies that basin extension should gradually decrease northwards along the rift system and reach a minimum in the North Tucano where NW-NNW strike-slip movement should have occurred without extension. Observed extension in the North Tucano Graben is larger than that predicted by the rotation model and internal deformation of the microplate is required to accomodate this extension. Although strike-slip deformation was probably important in the North Tucano, and caused widespread folding (Fig. 13), the basin also underwent important extension with large negative gravity anomalies indicating approximately 5 km of subsidence.

The progressive eastward narrowing of the Jatoba Basin is adequately explained by the rotation model. Further implication of the model are that the Vaza-Barris and the Itapicuru transfers are dextral transpressional, and the Mata-Catu transfer is dextral transtensional (Fig. 20). Existing seismic evidence does not contradict these movements. To compensate for the rift extension, compression should occur east of the rotation pole, and this may explain the reverse fault movements seen at the eastern end of the Pemambuco Lineament (Fig. 20). The amount of shortening at the eastern end of the lineament should be approximately 6 km, considering a 2” rigid rotation. Unfortunately, correlations are not possible across the lineament to permit an independent estimation of shortening. Discussion

of transfer faults

Strike-slip faults that trend obliquely to the main rift axis and can be shown to be active during rifting have been described by several authors (Bally, 1981; Gibbs, 1984; Bosworth, 1985; Lister et al., 1986) although they have mostly been shown on theoretical diagrams. The transfer faults within the Reconcavo-Tucano rift are particularly clear natural examples. Their role in the rift tectonics described here has several special characteristics and produces some unusual structures which deserve further mention. They can be summarized as follows: (1) They usually terminate against oblique normal faults, which may, or may not, exhibit a horizontal offset across the transfer (Fig. 21a). (2) They can terminate at, or continue across, basin margins (Fig. 21a). (3) There will generally be a change of throw, and perhaps dip, of the oblique normal faults at either side of their intersection with the transfer (Fig. 21b). (4) The movement sense may change along strike of the same transfer, and they may show a null point (N) where there is no displacement (e.g. Mata-Catu transfer, Fig. 21~). (5) The movement sense may change through time along the same fault segment.

68

d)

e)

Fig. 21. Characteristic structures and effects of transfer faults. a. Termination of transfer fault at highly oblique normal faults; map view. b. Change of throw and dip of normal faults either side of a transfer. c. Movement sense changes along strike of the same transfer: map view. d. Offset of depocenters by transfer which are parallel to the rift opening. e. Flip of depocenter across a transfer.

(6) The age of fault movement usually varies along strike. The whole fault will very rarely, if ever, be active at the same time. (7) They can accommodate initial offsets in the axes of maximum extension and subsidence, thus parallel transfer faults may show opposite apparent offsets (e.g. Mata-Catu and ItanagraArag&). If faults with opposite apparent offsets are vertical they should be sub-parallel to the rift opening direction (Fig. 216). (8) Basin asymmetry can change across the fault (e.g. the Vaza-Barris transfer, Fig. 21e). (9) Oblique normal faults near a transfer cannot be used in the same manner as normal faults in classic wrench tectonics to determine strike-slip sense. Maximum stress should be vertical in transfer zones, unlike wrench faults where it is subhorizontal. (10) They usually show rapid changes of throw along strike.

(11) They exhibit characteristics of strike-slip faults in seismic sections (Harding, 1985). (12) If the section is favourably oriented, oblique to the transfer, a cactus structure may be observed (e.g. Itapicuru transfer, Fig. 11). Conclusions

The Recancavo-TucanoJatoba rift is an aborted intra-~ntinen~l rift which opened during the Early Cretaceous South Atlantic rifting. The Rec&icavo-Tucano segment is believed to have opened obliquely to the overall N-S trend in a N30”-40 * W direction, about a rotation pole situated at 8”ll’S and 36”04’ W. The opening direction is roughly constrained by N30 o W transfer faults. Calculation using area balance of approximately 20% extension in the South Tucano and Recancavo grabens suggests that a 2” anticlockwise rotation of the East Brazilian Micro-

69

plate took place relative to the Sgo Francisco Craton. The Jatobi Graben is inte~reted as having opened in a N-S direction. Well-defined transfer faults were important in accommodating oblique opening. These faults show sudden changes of throw along strike, changes of movement sense along strike, offsetting of depocenters, and cactus and flower structures on seismic sections. “Cactus” is a new proposed term for a structure observed in oblique section cutting through opposing listric fans separated by a transfer fault. Basin architecture is strongly controlled by a complex mosaic of basement blocks, and is a particularly good example of nature’s way of finding the fracturing path of least resistance through the basement labyrinth. Basin asymmetry is strongly correlated with basement weakness. Depocenters are located close to the eastern margin of the Reconcavo, South and Central Tucano grabens, and are produced by N30 o E steep NWdipping normal faults developing parallel to basement shears and litholo~c~ contacts. The slightly flexured western margin has basement weaknesses which are unfavourably oriented for reactivation as down-to-the-basin normal faults. The flip of the depocenter to the western margin in the North Tucano Graben occurs across the Vaza-Barris Arch, which is a complex transfer zone controlled by Precambrian strike-slip faults separating stratigraphic terranes in the Sergipano Fold Belt. The E-W trend of the Jatoba Graben is controlled by the Pernambuco Lineament and subsidiary shear zone. Hence, the sigmoidal shape of the rift is believed to be controlled by basement anisotropies. Positive Bouguer anomalies (up to + 50 mGa1) flank the eastern side of the Tucano and Reconcavo grabens, but are not found on the western margin. The anomaly produced by the basin fill does not coincide with the maximum depocenter in the basin defined by seismic methods. These two important features of the gravity field strongly suggest that crustal thinning occurred below the rift. The calculated crustal profiles in the Tucano and Reconcavo Grabens show localized thinning up to 45% near to the eastern basin margin, and this geometry implies that a

large amount of lateral heat conduction probably occurred during the 20 Ma rifting period. This would cool the crust as it extended and may explain why no post-rift thermal subsidence phase exists.

The authors are grateful to PETROBRAS for permission to publish this paper. ID thanks this institution for kindly funding field work, and the Secretariat of Mines and Energy of Bahia for financial assistance, L.P. Ma~a~ta and J.A. Cupertino improved the text, and introduced us to the rift geology. References Bally, A.W., 1981. Atlantic type margins. In: A.W. Bally and A.B. Watts (Editors), Geology of Passive Continental Margins. Am. Assoc. Pet. Geol. Course Note Ser., 19: l-47. Bosworth, W., 1985. Geometry of propagating rifts. Nature, 316: 62.5-627. Brasil/DNPM, 1981. Mapa geologic0 do Brasil e da area ocegnica adjacent,, incluindo depositos minerais (Scale of 1 : 2500,000). Departamento National de Produ9aG Mineral, Brasilia. Christie-Blick, N. and Biddle, K.T., 1985. Deformation and basin formation along strike-slip faults. In: K.T. Biddle and N. Christie-Blick (Editors), Strike-Slip Deformation, Basin Formation and Sedimentation. Sot. Econ. Paleontol. Mineral., Spec. Publ., 37: l-34. Cochran, J.R., 1983. Effects of finite rifting times on the development of sedimentary basins. Earth Planet. Sci. Lett., 66: 289-302. Cohen, C., 1985. Role of fault rejuvenation in hydrogen accumulation and structural evolution of Rec&tcavo Basin, Northeast Brazil. Am. Assoc. Pet. Geol. Bull., 64: 65-76. Correia, E.G., 1965a. Contribui@o da refra@o sismica no delineamento do arcabouqo estrutural da Bacia de Jatoba. PETROBdS/RPBA, Salvador, internal rep. Correia, E.G., 1965b. Esqueleto tecti%rico da Bacia do Tucano Norte, Segundo OS dados da refra(;go &mica. PETROBRAS/RPBA, Salvador, internal rep. Ghignone, J.I., 1979. Geologia dos sedimentos fanerozoicos do Estado da Bahia. In: H.A.V. Inda (Editor), Geologia e Recursos Minerais do Estado da Bahia. Secretariat of Mines and Energy Basic Papers, 2: 24-117. Ghignone, J.I. and Andrade, G., 1970. General geology and major oil fields of Recancavo Basin, Brazil. In: M.T. Halbouty (Editor), Geology of Giant Petroleum Fields. Am. Assoc. Pet. Geol., 14: 337-358. Gibbs, A.D., 1983. Balanced cross-section construction from

70

seismic sections

in areas of extensional

tectonics.

J. Struct.

Geol. 5: 152-160. Gibbs,

A.D.,

Harding,

evolution

J. Geol. Sot. London,

of extensional

positive

flower structures,

structural

basin

141: 609-620.

T.P., 1985. Seismic characteristics

of negative

positive

inversion.

and identification flower structures

Am. Assoc.

Pet. Geol.

G.S., Etheridge, faulting

margins.

Geology,

M.A. and Symonds, and the evolution

do Recbncavo

and

rep.

Bull..

Ramsay.

P.A., 1986. De-

of passive

continental

cisalhante

na evolu@o

Rev.

Bras.

do rift do

GeociCnc.,

15:

287-292. E.J., 1987. Aspectos

do Reconcavo

e Tucano

Ci&ncia-T&mica-Petrbleo Netto,

A.S.T.,

Reconcavo. Netto.

A.S.T.

rift-valley 15: 97-102.

of 1: 250,000).

1985. ArcabouQo

e Tucano

das

bacias

do

Internal

estrutural

das bacias

Sul (Scale of 1: 250,000).

J.I., 1980. Shear zone geometry:

da evolu@o Sul, Bahia,

Congr.

Brasil.

das bacias

P.. Milani,

Lobo,

A.P.,

do

J.J., 1985. 0 preenchimento

do Reconcavo.

Rev. Bras. Geoci&nc.,

How

oil reserves

gravity

Internal

a review. J. Struct.

distribution. J.L. and

computations

Geophys.

Res., 64: 49959.

Tucano-Gabon

and

basin system.

M.O.

lithosphere. and

duced by oblique

Bott,

bodies

M.H.P.,

zone.

J.

1986. Crustal

and formation

of

322: 629-632.

sense normal

simple shear of the Sci., 22: 108-125.

1986.

Tectonophysics,

with ap-

fracture

rifting

Nature,

J.R.,

and

affects

M., 1959. Rapid

submarine

Can. J. Earth

Jamison, rifting.

J.C.J.

rifting

Oil Gas J., 83: 107-113.

Landisman,

South Atlantic

B., 1985. Uniform

continental Withjack,

G.D.

during

Atlantic

for two-dimensional

to the Mendocino

N., Kamer,

M.C., Concei+o.

South

plication

Wemicke, do

E.J., Lana,

1985.

M., Worzel.

detachment

18, 61 pp.

da fase rift na Bacia

Talwani,

Ussami.

PETROBI&

Bras. Geol. 30th, Recife, pp. 506-517.

Oliveira,

na Bacia

tectonica

Ser., Rio de Janeiro,

1978. A implanta@o and

Szatmari,

Brazilian

14: 246-250.

E.J., 1985. Tectanica

Reconcavo-Tucano-Jatoba. Milani,

Bouguer

(Scale

Geol. 2: 83-99.

tachment Milani,

Mapa

rep. PETROBRAS/DEXBA,

69: 582-600. Lister,

1970.

RecBncavo-Tucano-Jatoba

1984. Structural

margins.

PETROBRblS/RPBA,

Deformation 126: 99-124.

pro-