Caribbean Geology

Home , Cuche Formation, Fyzabad, Geology of Barbados, Geology of Cuba, Globorotalia menardii, Peroniceras

Caribbean Geology

An Introduction

Edited by

Stephen K. Donovan

and Trevor A. Jackson

Published by The University of the West Indies Publishers' Association (UWIPA) P.O. Box 42, Mona, Kingston 7, Jamaica W.I. Cover design by Annika C. Lewinson ISBN 976-41-0033-3

Printed in Jamaica

Contents

Introduction...... 1 STEPHEN K. DONOVAN and TREVOR A. JACKSON Chapter 1: Geologic Provinces of the Caribbean Region ...... 3 GRENVILLE DRAPER, TREVOR A. JACKSON and STEPHEN K. DONOVAN Chapter 2: Evolution of the Gulf of Mexico and the Caribbean...... 13 JAMES L. PINDELL Chapter 3: The Caribbean Sea Floor...... 41 THOMAS W. DONNELLY Chapter 4: Cuba ...... 65 GRENVILLE DRAPER and J. ANTONIO B ARROS Chapter 5: The Cayman Islands...... 87 BRIAN JONES Chapter 6: Jamaica ……………………………………………………………………………………. .1ll EDWARD ROBINSON Chapter 7: Hispaniola ...... 129 GRENVILLE DRAPER, PAUL MANN and JOHN F. LEWIS Chapter 8: Puerto Rico and the Virgin Islands ...... 151 DAVID K. LARUE Chapter 9: The Lesser Antilles ...... 167 GEOFFREY WADGE Chapter 10: Barbados and the Lesser Antilles Forearc...... 179 ROBERT C. SPEED Chapter 11: Tobago...... 193 TREVOR A. JACKSON and STEPHEN K. DONOVAN Chapter 12: Trinidad …………………………………………………………………………………...209 STEPHEN K. DONOVAN Chapter 13: Northern South America...... 229 STEPHEN K. DONOVAN Chapter 14: The Netherlands and Venezuelan Antilles ...... 249 TREVOR A. JACKSON and EDWARD ROBINSON Chapter 15: Northern Central America...... 265 BURKE BURKART Index ...... 285

iii

Introduction

STEPHEN K. DONOVAN and TREVOR A. JACKSON

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

THE LITERATURE of the geology of the Caribbean region ous reviewers who have given their time and energies in is widely dispersed through a number of primary sources. helping us to assess the various contributions to this volume. Apart from numerous research papers in international and In alphabetical order, we acknowledge the contributions of regional scientific journals, there are also the transactions I.E. Case (U.S. Geological Survey), T.W. Donnelly (State that have arisen from the various Caribbean, Latin American University of New York at Binghamton), J.F. Dewey (Uni- and Central American Geological Conferences (for refer- versity of Oxford), G. Draper (Florida International Univer - ences, see Draper and Dengo), plus various smaller, often sity), N.T. Edgar (U.S. Geological Survey), G.S. Home more specialized meetings, particularly in Jamaica and (Wesleyan University, Connecticut), V. Hunter (Laso Inc., Trinidad. To this burgeoning list can be added various Florida), B. Jones (University of Alberta), R.D. Liska review volumes, newsletters and unpublished reports. (Houston, Texas), P. Mann (University of Texas at Austin), Various general works have been published which re- J.D. Mather (Royal Holloway and Bedford New College, view this enormous literature, the most recent examples London), F. Nagle (University of Miami), R.K. Pickerill being edited by Nairn and Stehli (now almost 20 years old) (University of New Brunswick, Fredericton), M.J. Roobol and Dengo and Case1. These volumes are generally excel- (Directorate General of Mineral Resources, Saudi Arabia), lent, but are intended mainly as specialist references and are E. Robinson (U.W.L, Mona), K. Rodrigues (Trinidad and generally too expensive, and often too advanced or special- Tobago Oil Company Limited), R. Shagam (Rider College, ized, for student readers. This is particularly unfortunate for New Jersey), P.W. Skelton (The Open University), A.W. any undergraduate taking an advanced course in Caribbean Snoke (University of Wyoming at Laramie), R. Torrini, Jr. geology or, for that matter, any new graduate student start- (Woodward-Clyde Consultants), G. Wadge (University of ing research in the region. Caribbean Geology: An Intro- Reading) and G.K. Westbrook (University of Birmingham). duction has been produced to help fill the need for a cheap, T.A.J. also acted as a reviewer for two chapters. but comprehensive, text on Caribbean geology. The 15 Thanks, too, to Annika Lewinson and Annie Paul of the chapters have been written by the editors and a group of University of the West Indies Publishers' Association, who invited authors who are experts in particular aspects of the undertook the daunting task of typesetting this volume. Our geology of the region. While we have attempted to be as fellow authors endured entreaties by letter, fax, cable and comprehensive as possible, it is hoped that the text is pitched telephone, as well as patiently suffering delays at the Mona at a level that is both intelligible and informative to the end, but we trust that any feeling of persecution is forgotten student as well as the expert. now that the book is published. The Department of Geology We are pleased to acknowledge the financial support at UWI provided invaluable logistic support. that has made publication of this book possible. Printing was financed by a repayable grant from the Research and Publi- cations Fund of the University of the West Indies (UWI). REFERENCES Typesetting was supported by the Research and Publica- 1 tions Fund of Mona Campus, UWI. Dengo, G. & Case, J.E. (eds). 1990. The Geology of North We also make a particular point of thanking the numer - America, Volume H, The Caribbean Region. Geologi-

Introduction

cal Society of America, Boulder, Colorado, 528 pp. America, Boulder, Colorado. 2Draper, G.& Dengo, G. 1990. History of geological inves- 3Nairn, A.E.M. & Stehli, F.G. (eds). 1976. The Ocean tigation in the Caribbean region: in Dengo, G. & Case, Basins and Margins. 3. The Gulf of Mexico and the J.E. (eds), The Geology of North America, Volume H, Caribbean. Plenum, New York, 706 pp. The Caribbean Region, 1-14. Geological Society of

CHAPTER 1

Geologic Provinces of the Caribbean Region

GRENVILLE DRAPER1, TREVOR A. JACKSON2 and STEPHEN K. DONOVAN2

1Department of Geology, Florida International University, Miami, Florida 33199, U.SA. 2Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION South American continent. The western boundary comprises Central America and the Isthmus of Panama, and the THE CARIBBEAN is a geologically complex region that eastern limits are defined by the Lesser Antilles archipelago. displays a variety of plate boundary interactions including Within these boundaries there are several deeper water subduction in the Lesser Antilles and Central America, regions; the Yucatan Basin, the Cayman Trough, the Colom- transcurrent (strike-slip) motions on the northern and southern bian Basin, the Venezuelan Basin and the Grenada Basin. boundaries, and sea floor spreading in the Cayman These are separated by several more or less linear ridges and Trough. The central Caribbean is a lithospheric plate con- rises; by the Cayman Ridge, the Nicaraguan Rise, the Beata sisting mainly of an anomalously thick, oceanic plateau Ridge and the Aves Ridge, respectively. The physiographic situated between two major continental regions and therein units correspond in part to the different crustal provinces lies its geological importance. Classic studies of the Alps, that make up the Caribbean and in part to the active tectonic Himalayas and Appalachians have documented the effects elements that make up the present Caribbean Plate. of major continent-continent collisions. The Caribbean provides the opportunity to study the nature of the geological evolution of island arcs, and the tectonic interaction between PRESENT PLATE CONFIGURATION anomalously thick oceanic crust and continental crust. The purpose of this chapter is to introduce the physiog- The location and nature of plate boundaries in the Carib- raphy and geology of the Caribbean region. Although de- bean, as elsewhere, are determined by the location of earth- tailed analysis of tectonostratigraphic terranes has been quake hypocentres; by use of the sense of slip from first published previously11, in the present account we attempt to motion studies on seismogenic faults; from detailed outline the features of the major geologic provinces that bathymetric, magnetic and seismic profiling studies of ma- make up the Caribbean and to provide a framework for the rine areas; and from detailed mapping of recent, on-land more detailed descriptions which follow in this volume. structures if the plate boundary happens to be exposed onshore. These studies show that the boundaries of the Carib- PHYSIOGRAPHIC PROVINCES bean Plate (Fig. 1.1 A), as defined by the distribution of earthquake epicentres53, run approximately from Guate- The Caribbean region is comprised of several major marine mala along the trend of the Cayman Trough, through His- and terrestrial physiographic and geologic provinces, the paniola and Puerto Rico, south through the Lesser Antilles, geographic relationships of which are illustrated in Figure and along the northern South America continental margin 1.1. Geographically and bathymetrically, the Caribbean Sea (although this boundary is poorly defined between Trinidad is bound to the north by the Gulf of Mexico, the Yucatan and the Meridional Andes) and the west coast of Central Platform, the Florida-Bahamas Platform and the Puerto America. First motion solutions7,28 indicate left-lateral Rico Trench, and to the south by the northern part of the strike slip at the northern boundary and right-lateral strike

Geologic Provinces of the Caribbean Region slip at the southern boundary, indicating that these are left important geophysical features of the Caribbean Sea floor. lateral and right lateral transform boundaries, respectively. The B" horizon marks the boundary between igneous sills Thrust fault solutions, typical of the upper part of convergent and overlying Upper Turonian to Coniacian sedimentary plate boundaries, are found in both the western and eastern strata. This reflector has been traced from the lower Nicara- margins of the plate. The depth of the hypocentres and their guan Rise eastwards through the Venezuelan Basin . The position relative to island arc volcanoes indicates Wadati- A" horizon is considered to mark the boundary between Benioff Zones dipping eastward beneath Central America Lower to Middle Eocene oozes and chalks, and underlying and westward beneath the Lesser Antilles. Detailed marine Upper Cretaceous chertiferous limestones. This reflector and terrestrial studies have considerably refined this general extends from the lower Nicaraguan Rise in the west to the picture, although there are considerable differences of opin- Grenada Basin in the east. Borehole data from Deep Sea ion about the details of the present direction and rate of Drilling Project Leg 15 indicates that the B" horizon consists movement of the Caribbean Plate relative to its neighbours. of the uppermost layers of a large oceanic basalt plateau with a The Caribbean Plate is moving eastwards with respect to crustal thickness of between 15 and 20 km. This plateau both North and South America31 at a rate of about 1 to 2 was produced by a significant oceanic flood basalt event that cm yr-1. The northern and southern boundaries of the plate occurred during the late Cretaceous14 (Donnelly, Chapter 3, are thus transform fault systems dominated by left-lateral herein). and right-lateral strike-slip motions, respectively. Unlike

transform fault systems in oceanic crust, where the move- Colombian Basin ment is accommodated in single, discrete fault zones, the The Colombian Basin is defined by the Hess Escarp- movements in the Caribbean are distributed on several ac- ment to the north, and the continental margin of Panama and tive fault zones to produce broad, active seismic zones about Colombia to the south. The Colombian Plain, extending to 200 km wide. As it is difficult to pinpoint the precise plate depths of 4000 to 4400 m, is the largest abyssal plain in the boundary, the north and south Caribbean Plate boundaries 8,28 Caribbean region and is located in the northeast part of the are best characterised as plate boundary zones . basin. This plain extends north to Hispaniola, south to the The Motagua, Polochic and other fault zones form Magdelana Fan and east to the southern corner of the Beata the eastern extension of the Northern Caribbean Plate Ridge22. Both the Magdelana and the Panama-Costa Rica Boundary Zone in Central America. A left lateral step- Fans introduce significant quantities of sediment into the over in the boundary between the Caribbean and North basin along its southern and western edges. The North America has resulted in a crustal-scale pull-apart basin, Panama and South Caribbean Deformed Belts are under - the Cayman Trough, in which a 100 km long spreading thrust margins to this basin. ridge segment has been produced. This ridge is bounded by two extensive transform faults, the Swan Island Transform Fault and the Oriente (previously Bartlett) Beata Ridge Transform Fault. East of Cuba, left lateral displacement The Beata Ridge is a structural high that extends south- may be accomodated on several fault zones in northern west from Cape Beata, Hispaniola, for about 400 km. The Hispaniola and offshore. Left lateral displacement has also ridge has a relief of about 2000 m and is comprised of a been documented south of the Cayman Trough in Jamaica series of north-south trending subsidiary ridges which be- 8,29,30 come less pronounced towards the south, where it converges and southern Hispaniola , and forms the southern 11,26 boundary of a microplate 44. on the South Caribbean Deformed Belt . Initial uplift of the The eastward movement of the Caribbean Plate has ridge occurred during the late Cretaceous and coincided with structural disturbances that affected the northern Colombian resulted in subduction of the Atlantic Ocean crust under 23 the eastern margin of the Caribbean, producing the Lesser Basin and the Hess Escarpment . Subsequent tectonic Antilles island arc system. Eastward motions of the Pacific events have led to the tilting of the ridge and to and Cocos Plates with respect to the Caribbean and North deformation along its southern margin.

Amer ica are equally rapid, which has resulted in subduction of these plates beneath the western margin of Venezuelan Basin the Caribbean, that is, under Central America. The Venezuelan Basin is the deepest and largest of the Caribbean basins. The interior of the basin includes less than 200 m of relief, having been 'smoothed' by the accumulated GEOLOGIC PROVINCES —CARIBBEAN SEA sediments. The basin is deepest at its northern (Muertos Trough) and southern (Venezuelan Plain) boundaries, Seismic reflectors A”and B” where it converges with the North and South Caribbean Persistent seismic reflector horizons A" and B" are Deformed Belts, respectively11. Most of the sediment in the

G. DRAPER, T.A. JACKSON and S.K. DONOVAN

Figure 1.1. (A) Map of the Caribbean region showing the relative positions of plates, physiographic regions and major islands (redrawnafter Jackson24). Direction of subduction shown by solid triangles. (B) Geologic provinces of the Caribbean region, as defined in the present chapter (simplified after Case and Dengo10; Case et al.12). Key: AP=Anegada Passage; AR=Aves Ridge; BeR=Beata Ridge; BP=Bahamas Platform; BR-Barbados Ridge and Lesser Antilles Deformed Belt; C=Cuba; C A=Colombian Andes; CB=Chortis Block; ChB=Choco Block; CO=Cuban Orogenic Belt; CoB=Colombian Basin; CT=Cayman Trough; CtB=Chorotega Block; EPFZ=E1 Pilar Fault Zone; GA=Greater Antilles; GAOB=Greater Antilles Orogenic Belt; GB=Grenada Basin; GM=Gulf of Mexico; H=Hispaniola (Haiti+Dominican Republic); J=Jamaica; LA=Lesser Antilles; MPFZ=Motagua-Polochic Fault Zone; NP=Nazca Plate; NPD=North Panama Deformed Belt; NR=Nicaraguan Rise; OTF=Oriente Transform Fault; PR=Puerto Rico; SCD=South Caribbean Deformed Belt; SITF=Swan Island Transform Fault; VB=Venezuelan Basin; VBo=Venezuelan Borderland; YB=Yucatan/Maya Block; YBa=Yucatan Basin.

Geologic Provinces of the Caribbean Region basin was derived from the eastern margin, marked by the Chorotega and Choco Blocks Aves Apron that extends westwards from the Aves Ridge. Geographically, this province comprises Costa Rica, Panama and northwestern Colombia. The Choco Block, west of and overridden by the Cordillera Occidental of the GEOLOGIC PROVINCES—CENTRAL AMERICA Colombian Andes, comprises a sequence of uplifted Upper Cretaceous to Paleogene oceanic crust and magmatic arc Yucatan/Maya Block rocks. The Choco Block abuts deep forearc basins to the The Yucatan, or Maya, Block is located on the North west containing up to 10,000 m of pelagic, turbiditic and American Plate and is separated from the Chortis Block by marginal marine sediments and sedimentary rocks 11. East- the Motagua-Polochic fault system. The (pre-Carbonifer- ern Panama (Choco Block) is a raised block with a basement ous) basement of the Yucatan Block is exposed in its southern of late Cretaceous or older oceanic crust, topped by seismic extremity, near the Motagua suture, and occurs in various reflector BM (see above), and overlain by Upper Cretaceous wells in the subsurface. It is composed of schists, marbles, pelagic sedimentary rocks9. Panama became attached to quartzites and granitoids of unknown age. These rocks are South America and nuclear Central America during the late unconformably overlain by Upper Carboniferous to Permian Miocene or Pliocene13. sedimentary and volcanic rocks 15. The Chorotega Block is essentially the northern exten- The Palaeozoic sedimentary rocks are unconformably sion of the Choco terrane and comprises a series of belts overlain by a Jurassic 'red bed' sequence (Todos Santos parallel to the Pacific coast developed by subduction of the Group) and thick, Cretaceous dolomitic limestones. In the Cocos Plate and subsequent accretion of terranes on the southern part of the block, Upper Cretaceous to Tertiary western seaboard. Reversal and repetition of the forearc olistostromes and immature sandstones of the Sepur Group ridge and basin occur in the east of Panama1 . The overlying overlie the carbonate platform rocks. The Chicxulub Crater in Middle America volcanic province is a northwest-southeast the middle of the Yucatan carbonate platform is the trending belt in western Panama consisting of Miocene to probable candidate for the Cretaceous/Tertiary boundary Holocene calc-alkaline volcanics and related deposits. This impact crater21. volcanism is related to subduction at the Middle America Trench11. Chortis Block The northern boundary of the Chortis Block on the GEOLOGIC PROVINCES—NORTHERN Caribbean Plate is defined by the Motagua-Polochic fault CARIBBEAN system (at present, an active strike-slip fault zone, but this was previously a suture zone formed by the late Cretaceous Gulf of Mexico collision of the Chortis and Yucatan Blocks), which is also The northern margin of the GuIf of Mexico is underlain the boundary between the Caribbean and North America by a broad zone of stretched and thinned continental crust, Plates. The southwestern boundary is the Middle America as is the southern part (see Maya/Yucatan Block, above). Trench, which separates the block from the Cocos Plate. The central part of the GuIf of Mexico is underlain by However, the southern and eastern boundaries are less well Upper Jurassic to Lower Cretaceous oceanic crust. This defined. structure is the result of the rifting of the Maya/Yucatan The basement of the Chortis Block consists of pre- Block from North America38,39,46, which resulted from Mesozoic (probably Palaeozoic) metamorphic rocks and approximately northwest-southeast continental extension associated Mesozoic plutons which outcrop in the northern that took place from the Triassic to the late middle and central parts of the block. The Mesozoic sequence is Jurassic. This was followed by sea floor spreading until the generally similar to that of the southern part of the Yucatan earliest Cretaceous. Block to the north, but less well documented15. A Jurassic to The continental basement on the northern and southern Lower Cretaceous 'red bed' sequence (correlated with the margins of the Gulf of Mexico is overlain by Triassic and Todos Santos Group of the Yucatan Block) overlies the Jurassic 'red beds' and Jurassic evaporites (Louann and basement rocks. These are in turn overlain by massive Campeche provinces, respectively). Uppermost Jurassic Lower Cretaceous limestones. A major Upper Cretaceous sedimentary rocks comprise shallow-water limestones on 'red bed' sequence (Valle de Angeles Formation) sits on the margins of the Gulf, with deep-water carbonate facies in these limestones. A major unconformity separates the the central regions. A similar pattern persisted through the Mesozoic sequence from extensive Cenozoic volcanic de- Cretaceous and produced thick carbonate sequences. These posits. limestones are overlain in the western and central Gulf of Mexico by terrigenous clastic sedimentary rocks derived as

G. DRAPER, T. A. JACKSON and S.K. DONOVAN a result of late Cretaceous orogenic uplift in western North deformation occurred earlier in the west. The islands consist America and Mexico. of a Jurassic oceanic basement (exposed in central His- paniola and southwest Puerto Rico) overlain by Lower Florida and Bahamas platforms Cretaceous (Aptian-Albian or possibly older) to Paleogene The Florida and Bahamas carbonate platforms (here island arc deposits (volcanic and epiclastic deposits with taken to include the regions underlying the Turks and Caicos associated immature clastic and carbonate sedimentary Islands) lie to the north of both the Cuban Orogenic Belt and rocks). Although there is some evidence for Cretaceous the Hispaniola segment of the Greater Antilles Orogenic deformation events16,17,33,34, which may have a similar age Belt The Florida and Bahamas platforms consist of a con- to those in Cuba, the major tectonic deformation in these tinuous sequence of Middle Jurassic to Recent carbonate islands was usually later. Late Cretaceous to early Paleogene sedimentary rocks which are over 6,000 to 7,000 m thick in deformation was associated with an oblique collision of the 48 southern Florida and over 10,000 m thick in the Bahamas . Greater Antilles arc with the Florida-Bahamas platform, but The accumulation of these limestones resulted from the which did not appear to produce the extensive thrusting subsidence accompanying the rifting that formed the Atlan- otherwise seen in Cuba. In contrast, in Jamaica (see Nicara- tic Ocean and Gulf of Mexico. In Florida, the northward guan Rise, below), the early Paleogene was a period of thinning carbonate accumulations unconformably overlie crustal extension and resulted in deposition of rift facies Triassic to Lower Jurassic arkoses and volcaniclastic sedi- sediments. From the post-Oligocene to the present, the mentary rocks, which in turn rest on Palaeozoic basement. islands have experienced another major orogenic phase due The situation is similar in the western Bahamas, under the to sinistral transpression caused by the eastward motion of Great Bahama Bank, but east of New Providence island the the Caribbean Plate relative to North America. These 48 basement consists of Jurassic oceanic crust . transcurrent movements produced a series of strike-slip related, clastic -filled basins. In adjacent areas, moderate Cuban Orogenic Belt subsidence coupled with eustatic sea level changes pro- Western and central Cuba form a major orogenic belt, duced carbonate build-ups. characterized by northwardly-directed thrusting of Creta- ceous island arc volcanic rocks, with associated oceanic Nicaraguan Rise crust, over a sequence of continental shelf to slope, Jurassic The Nicaraguan Rise extends northeastwards from to Lower Cretaceous limestones and mature clastic sedi- Honduras and Nicaragua in Central America to Jamaica and mentary rocks. It was previously thought that this orogenic southern Haiti2. It is bounded on the northern edge by the belt resulted from the collision of an island arc with the Cayman Trough and along the southern margin by the Florida-Bahama continental margin in the late Cretaceous northeast-southwest trending Hess Escarpment. The Nica- 20 to early Tertiary . The Campanian ages of olistostromes raguan Rise is a broad, topographically complex feature of deposited at the front of advancing thrust sheets, and Cam- shallow to intermediate depth (0-3000 m) along which there panian metamorphism of continental margin sedimentary is an upper (less than 1200 m water depth) and lower (greater rocks in southern Cuba indicate that the Cuban orogeny than 1200 m water depth) rise22. The lower Nicaraguan Rise comprised a middle Cretaceous, and an early Cenozoic, is separated from the upper part by the Pedro Bank escarp- 41 orogenic events (Draper and Barros, chapter 4, herein). ment or Pedro Bank Fracture Zone11 and from the Colom- Cuba is the only region in the Greater Antilles where bian Basin by the Hess Escarpment. The lower, or Precambrian age rocks occur. Grenville age (approximately southern11, Nicaraguan Rise is comprised of a series of 1,000 Ma) metamorphic rocks outcrop in Las Villas prov- faults, troughs and volcanoes. This is particularly evident 42,50 ince in north central Cuba . These rocks may represent an along the northeast and southwest margins, where there are exposed fragment of the basement underlying the sedi- prominent rifts (the Morant and San Andres Troughs, re- mentary rocks of the continetal margin. spectively). Closely associated with these troughs is a series Southeastern Cuba contains rocks formed in a Paleo- of seamounts and islands formed from late Cenozoic vol- cene island arc which has a geological history distinct from canic rocks57. western and central Cuba. Lower rise strata vary in thickness between 500 m and 1000 m, whereas the upper rise, which is mainly a carbonate Greater Antilles Orogenic Belt platform, is underlain by over 5000 m of strata11. Emergent The Greater Antilles orogenic belt comprises His- portions of the upper rise include Jamaica, as well as several paniola, Puerto Rico, the Virgin Islands and southeastern carbonate banks to the south of the island. The known Cuba. This province differs from that of the orogenic belt of stratigraphy of the upper rise, determined from various western and central Cuba in style of deformation, although sources 23,36,37 (Robinson, chapter 6, herein), suggests that

Geologic Provinces of the Caribbean Region

a basement of Upper Jurassic(?) to Lower Cretaceous oce- nated by the Cayman Ridge, a subsided volcanic arc devel- anic oust (with continental oust in the west) is overlain by oped on pre-Cenozoic oceanic(?) crust. In the east this predominatly stratified Upper Cretaceous volcanic rocks crustal block dips northeast beneath the Cuban margin. and Tertiary limestones. Geophysical data show the maxi- 1,2 mum thickness of the rise to be about 22 km . GEOLOGIC PROVINCES —EASTERN CARIBBEAN Cayman Trough The Cayman Trough is approximately 1600 km in Lesser Antilles length, 120 km in width and 5 km deep. It comprises a floor The Lesser Antilles volcanic arc is comprised of a series of thin oceanic crust (less than 7 km) partly overlain by a of islands stretching from Grenada in the south to the 37 veneer of younger sediments . The trough extends west- Anegada Passage in the north, a distance of 850 km. It is wards from the Windward Passage to the Gulf of Honduras separated from the Barbados Ridge in the south by the and separates the Cayman Ridge from the Nicaraguan Rise. Tobago Trough, a forearc basin, and from the Aves Ridge The Cayman Ridge may be a fragment of the Nicaraguan by the Grenada Basin. Rise which became separated by the opening of the Cayman The area has been described as a double arc -32 Trough. system" In the southern half of the chain the two arcs The approximately east-west trending Oriente and are superimposed on one another to form the islands of Swan Island Transform Faults are connected by the north- Grenada, the Grenadines, St. Vincent, St. Lucia and south trending Mid-Cayman Rise. The last is the site of Martinique. These islands contain volcanic and east-west seafloor spreading and hence the trough is essen- sedimentary rocks that range in age from the middle 30 tially a crustal-scale pull-apart basin . Therefore, the Cay- Eocene to the Holocene35. North of Martinique the arc man Trough is an important tectonic feature, as the rate of bifurcates into an older outer ridge and a younger inner spreading on the Mid-Cayman Rise must be equal to the ridge (the Limestone Caribees, inactive for the past 28 relative rate of movement of the Caribbean, with respect to Myr, and Volcanic Caribees, respectively). The Volcanic the North American Plate. Caribees have a history of late Tertiary and Quaternary The timing of the opening of the Cayman Trough volcanism (Wadge, chapter 9, herein). 27 remains unresolved. MacDonald and Holcombe con- Most of the volcanic activity in the Lesser Antilles is tended that the Cayman Trough is no older than Miocene, subaerial, as recorded by the eruptions in St Kitts, Guade- 45 whereas Rosencrantz and Sclater recognised magnetic loupe, Martinique, St. Lucia and St. Vincent during historic anomalies that trace the opening back to the mid Eocene. times49. Sea-going surveys6 have shown that the only active This also has implications regarding spreading rates. Mac - submarine volcano in the region is Kick-'em-Jenny, which 27 Donald and Holcombe considered that spreading rates is located just north of Grenada in the Grenadines. were 2 cm yr-1 for 0 to 2.4 Ma and 4 cm yr"1 for 2.4 to 6 Ma, whereas Rosencrantz and Sclater postulated rates of 1.5 Barbados Ridge -1 -1 cm yr for 0 to 30 Ma and 3 cm yr before then. Recent The Barbados Ridge is a forearc ridge that emerges 18,44 GLORIA and SeaMARC II sidescan mapping has re- above sea level at Barbados, an island capped with Pleisto- vealed a more complex spreading history punctuated by cene limestones and underlain by deformed Tertiary sedi- intervals of rise jumping. mentary rocks (Speed, Chapter 10, herein). The ridge is divided into an inner (=arcward) zone and an outer (=ocean- Yucatan Basin ward) region51, and forms part of the western margin of the The Yucatan Basin is bounded to the south by the Lesser Antilles accretionary prism25, which is over 300 km Cayman Ridge, to the west by the Yucatan Peninsula of wide. Mexico, and to the north by western and central Cuba The inner zone of the Barbados Ridge consists of rocks 43 Rosencrantz divided the basin and its borderlands into and structures similar to those of the Paleogene basal com- nine domains based on seismic reflection studies and surface plex of Barbados, including turbidites, olistostromes and topography. These domains occur on three distinct types or volcaniclastic sedimentary rocks. The thickness of this low blocks of crust. In the west, the eastern shelf of the Yucatan density rock sequence may be as much as 20 km. Rocks of Peninsula is characterized by northnortheast-southsouth- similar compos ition occur in the outer (eastern) region of west trending extensional faults and grabens. This Yucatan the Barbados Ridge, but here consist of recently accreted, borderland is flanked by a rectangular deep that occupies the fault-bounded packets51. western third of the basin. The floor of the eastern two thirds Accretion may have commenced in the Eocene during of the basin is topographically heterogeneous, but is domi- the early growth of the Lesser Antilles island arc. The

G. DRAPER, T.A. JACKSON and S.K. DONOVAN

differing structures of both the inner and outer zones may Plate. The emergent parts of the belt comprise the Caribbean reflect two discrete accretionary events. Speed51 suggested Mountain System of northern Venezuela and the southern that the inner zone developed above a lithospheric slab Antllean island chain of Aruba, Bonaire and Curaçao. The descending to the northwest, while the outer zone was central part of this island chain is separated from the main- formed above a slab descending to the west. From the land by the 2000 m deep Bonaire basin. The Curaçao Ridge commencement of accretion (about 50 Ma) to the present is a bathymetric high located north of the Netherlands An- the Barbados Ridge has been rising as a result of the growth, tilles and separated from it by the Los Roques Trough. This thickening and backthrusting of the accretionary com- area represents a zone of intense deformation and accretion, 25,54 plex . and is an eastward extension of the South Caribbean De- formed Belt that contains about 10 km of Paleogene(?) and 11,23,26 Aves Ridge Neogene pelagic and turbiditic deposits . The La Or- The Aves Ridge is located 200 km west of the Lesser chila Basin, a northwest-southeast trending graben, sepa- Antilles arc and extends in a north-south direction for about rates the Venezuelan Antillean islands of Los Aves, Los 500 km. It is a plateau between 50 and 150 km in width that Roques and La Qrchila from La Blanquilla, Los Hermanos, includes several steep-sided, north-south trending pedestals, Margarita, Los Frailes and Los Testigos to the east. one of which rises above sea level to form Aves Island56. The South Caribbean Island Chain comprises islands of The rocks underlying the ridge comprise a basement of the Netherlands and Venezuelan Antilles which extend from Cretaceous to Paleocene basalts, andesites and granites, that Los Monjes in the west, eastward to Los Testigos. These are overlain by about 1500 m of pelagic and shallow-water islands are located on an east-west structural high and ap- 3 Tertiary sedimentary rocks 11. The crustal character and pear to be genetically related to one another (Jackson and composition of the rocks of the Aves Ridge suggest that the Robinson, Chapter 14, herein). The Netherlands and Vene- area represents the site of an extinct magmatic arc5,23,40,52. zuelan Antilles consist of weakly metamorphosed Creta- ceous volcanic and sedimentary rocks which were intruded Grenada Basin by late Cretaceous granitoid bodies of various sizes, and are The Grenada Basin separates the southern portions of capped by late Cenozoic sedimentary rocks. the Aves Ridge from the Lesser Antilles arc. In the south the The Caribbean Mountain System is an east-west trend- basin attains depths of about 3000 m, but to the north the ing belt that extends from Sierra Nevada de Santa Marta in 4 water depth decreases, and the Aves Ridge and Grenada the west to the island of Tobago in the east . Six thrust- Basin merge into a single platform called the Saba Bank. bounded nappes outcrop discontinuously along the north Major lithospheric changes in the north and south of the coast of Venezuela These nappes consist of Cretaceous basin occur between latitudes 14°N and 15°N. To the south sequences that have been emplaced southward onto Paleo- the lithosphere is typical of a back-arc basin, composed of gene sedimentary rocks in a foreland basin setting. anomalously thick, two-layer oceanic crust similar to that of the nearby Venezuelan Basin40. This crust is overlain by Colombian Andes about 6 km of Tertiary volcaniclastic and pelagic sedimen- The major tectonic blocks included within the Colom- tary rocks 47. North of 15°N the basement is disturbed and bian Andes19 are, approximately from west to east, the is overlain by about 2 km of sediments and sedimentary Cordillera Occidental, the Cordillera Central, the Cordillera rocks11. The consensus of opinion is that the Grenada Basin Oriental, the Sierra Nevada de Santa Marta, the Sierra de is an intra-arc basin created by the splitting of a mature arc Perija and the Cordillera de Mérida, with associated sedi- during the early Paleogene into the Aves Ridge and the mentary basins (Donovan, chapter 13, herein). Lesser Antilles arc39. Alternately, the Grenada Basin may The Cordillera Occidental is a fault-bounded block be thinned forearc crust that became isolated during an consisting of a Mesozoic eugeosynclinal sequence devel- eastward migration of the subduction zone to its present oped on oceanic crust and intruded by Tertiary granitoid site23. plutons. It is separated from the Cordillera Central, which has a basement of continental crust, by the Romeral Fault GEOLOGIC PROVINCES-NORTHERN SOUTH Zone, which is characterised by a melange of oceanic and AMERICA continental fragments beneath a Tertiary cover sequence. The Cordillera Central is a polydeformed metamorphic Venezuelan borderland complex consisting of rocks from Precambnan to Creta- The Venezuelan borderland forms part of a broad oro- ceous, or uncertain, age55. A Precambnan to Palaeozoic genic belt of Mesozoic and Cenozoic rocks that mark the crystalline core includes a metamorphosed Lower Palaeo- boundary zone of the Caribbean with the South American zoic island arc sequence, and is overlain by Mesozoic to

Geologic Provinces of the Caribbean Region

Cenozoic marine and continental deposits that have been ern Caribbean Plate Boundary. Journal of Geophysical intruded by plutons. The Sierra Nevada de Santa Marta to Research, 83,375-386. the north, a pyramidal, fault-bounded massif in isostatic 9Case, J.E. 1974. Oceanic crust forms basement of eastern disequilibrium, includes a similar sequence to the Panama. Geological Society of America Bulletin, 85, Cordillera Central. The Cordillera Oriental, Sierra de 645-652. Perija and Cordillera de Merida (=Venezuelan Andes) 10Case, J.E. & Dengo, G. 1982. The Caribbean region: in have broadly similar stratigraphies and structures, with Palmer, A.R. (ed.), Perspectives in Regional Geological Precambrian to Palaeozoic crystalline basement overlain Synthesis: Planning for the Geology of North America. by Palaeozoic to Cenozoic, mostly continental Geological Society of America DNAG Special sedimentary and volcanic sequences. The Cordillera Publication, 1,163-170. Oriental has been autochthonous on nuclear South America 11Case, J.E., Holcombe, T.L. & Martin, R.G. 1984. Map of since the Jurassic. geologic provinces in the Caribbean region. Geological Society of America Memoir, 162, 1-30. ACKNOWLEDGEMENTS—We thank Paul Mann (University of Texas at 12Case, J.E., MacDonald, W.D. & Fox, P.J. 1990. Caribbean Austin) and Kirton Rodrigues (Trintoc, Pointe-a-Pierre) for making constructive review comments on this typescript crustal provinces: seismic and gravity evidence: in Dengo, G. & Case, J.E. (eds), The Geology of North America. Volume H. The Caribbean Region, 15-36. REFERENCES Geological Society of America, Boulder. 13Collins, L.S. & Coates, A.G. 1992. Timing and rates of 1 Arden, D.D., Jr. 1969. Geologic history of the Nicaraguan emergence of the northwestern Panama microplate: Rise. Transactions of the Gulf Coast Association of Caribbean effects of Cocos Ridge subduction? Geo- Geological Societies, 19, 295-309. logical Society of America Abstracts with Programs, 2 Arden, D.D., Jr. 1975. Geology of Jamaica and the Nicara- 24(7), A64. guan Rise: in Nairn, A.E.M. & Stehli, F.G. (eds), The 14Donnelly, T.W. 1973. Circum-Caribbean explosive vol- Ocean Basins and Margins. Volume 3. The Gulf of canic activity: evidence from Leg 15 sediments: in Mexico and the Caribbean, 617-661. Plenum, New Edgar, N.T. & Saunders, J. (eds), Initial Reports of the York. Deep Sea Drilling Project, 15, 969-988. 3 Beets, D.J., Maresch, W.V., Klaver, G.T., Mottana, A., 15Donnelly, T.W., Home, G.S., Finch, R.C. & Lopez-Ra- Bocchio, R., Beunk, F.F. & Monen, H.P. 1984. Mag- mos, E. 1990. Northern Central America: the Maya and matic rock series and high-pressure metamorphism as Chortis Blocks: in Dengo, G. & Case, J.E. (eds), The constraints on the tectonic history of the southern Car- Geology of North America. Volume H. The Caribbean ibbean. Geological Society of America Memoir, 162, Region, 37-76. Geological Society of America, Boul- 95-130. der. 4 Bellizzia, A. & Dengo, G. 1990. The Caribbean Mountain 16Draper, G. & Lewis, J.F. 1982. Petrology and structural system, northern South America: a summary: in Pengo, development of the Duarte complex, central Dominican G. & Case, J.E. (eds), The Geology of North America. Republic; a preliminary account and some tectonic Volume H. The Caribbean Region, 167-175. Geological implications: in Snow, W., Gil, N., Llinas, R., Ro- Society of America, Boulder. driguez-Torres, R., Seaward, M. & Tavares, I. (eds), 5 Bouysse, P., Andrieff, P., Richard, M., Baubron, J.C., Transactions of the Ninth Caribbean Geological Con- Mascle, A., Maury, R.C. & Westercamp, D. 1985. Aves ference, Santo Domingo, Dominican Republic, August Swell and northern Lesser Antilles ridge: rock-dredging 16th-20th, 1980,1, 53-64. results from ARCANTE 3 cruise: in Mascle, A. 17Draper, G. & Lewis, J.F. 1992. Metamoiphic belts in (ed.), Geodynamique des Caraibes, 65-76. Technip, central Hispaniola. Geological Society of America Special Paris. Paper, 262,29-45. 6 Bouysse, P. & Sigurdsson, H. 1982. The "Hodder Phe- 18Edgar, N.T., Dillon, W.P., Jacobs, C., Parsons, L.M., nomenon" of 1902: no active volcano off St. Lucia Scanlon, K.M. & Holcombe, T.L. 1990. Structure and (Lesser Antilles). Marine Geology, 50, 1-2,1129-1136. spreading history of the central Cayman Trough: in 7 Burke, K., Cooper, C., Dewey, J.F., Mann, P. & Pindell, La rue, D.K. & Draper, G. (eds), Transactions of the J.L. 1984. Caribbean tectonics and relative plate mo - Twelth Caribbean Geological Conference, St. Croix, tions. Geological Society of America Memoir, 162, U.S.V.L, August 7th-l1th, 1989,33-42. 31-63. 19Gansser, A. 1973. Facts and theories on the Andes. Journal 8 Burke, K., Grippi, J. & Sengor, A.M.C. 1980. Neogene of the Geological Society of London, 129,93-131. structures in Jamaica and the tectonic style of the North-

G. DRAPER, T. A. JACKSON and S.K. DONOVAN

20Gealey, W.K. 1980. Ophiolite obduction mechanisms: in Caribbean neotectonics: in Dengo, G. & Case, J.E. Ophiolites: Proceedings of the International Ophiolite (eds), The Geology of North America. Volume H. The Symposium, Cyprus, 1979, 243-247. Cyprus Geologi- Caribbean Region, 307-338. Geological Society of cal Survey Department, Nicosia. America, Boulder. 32 21Hildebrand, A.R., Penfield, G.T., Kring, D. A., Pilkington, Martin-Kaye, P.H.A. 1969. A summary of the geology of M., Camargo Z., A., Jacobsen, S.B. & Boynton, W.V. the Lesser Antilles. Overseas Geology and Mineral 1991. Chicxulub Crater: a possible Cretaceous/Tertiary Resources, 10, 172-206. 33 boundary impact crater in the Yucatan Peninsula, Mex- Mattson, P.H. 1960. Geology of the Mayaguez area, ico. Geology, 19, 867-871. Puerto Rico. Geological Society of America Bulletin, 22Holcombe, T.L. 1977. Caribbean bathymetry and sedi- 71, 319-362. 34 ments: in Weaver, J.D. (ed.), Geology, Geophysics and Mattson, P.H. & Pessagno, E.A., Jr. 1979. Jurassic and Resources of the Caribbean: Report of the IDOE Work - early Cretaceous radiolarians in Puerto Rican ophiolite shop on the Geology and Marine Geophysics of the - tectonic implications. Geology, 7,440-444. Caribbean region and its Resources, Kingston, Ja- 35Maury, R.C., Westbrook, G.K., Baker, P.E., Bouysse, P. maica, February 17-22, 1975,27-62. & Westercamp, D. 1990. Geology of the Lesser An- 23Holcombe, T.L., Ladd, J.W., Westbrook, G., Edgar, N.T. tilles: in Dengo, G. & Case, J.E. (eds), The Geology of & Bowland, C.L. 1990. Caribbean marine geology: North America. Volume H. The Caribbean Region, ridges and bas ins of the plate interior: in Dengo, G. & 141-166. Geological Society of America, Boulder. Case, I.E. (eds), The Geology of North America. Vol- 36Meyerhoff, A. A. & Kreig, E.A. 1977. Petroleum potential of ume H. The Caribbean Region, 231-260. Geological Jamaica. Special Report, Ministry of Mining and Society of America, Boulder. Natural Resources, Mines and Geology Division, Ja- 24Jackson, T. A. 1994. The marine geology and the maica, 131 pp. 37 non-living resources of the Caribbean Sea: an over- Perfitt, M.R. & Heezen, B.C. 1978. The geology and view. Caribbean Marine Studies, 2 (for 1991), 10-17. evolution of the Cayman Trench. Geological Society 25Ladd, J.W., Holcombe, T.L., Westbrook, G. & Edgar, of America Bulletin, 89, 1155-1174. N.T. 1990. Caribbean marine geology: active margins 38Pindell, J.L. 1985. Alleghenian reconstruction and the of the plate boundary: in Dengo, G. & Case, J.E. (eds), subsequent evolution of the Gulf of Mexico, Bahamas The Geology of North America. Volume H. The Carib and Proto-Caribbean Sea. Tectonics, 4,1-39. bean Region, 261-290. Geological Society of America, 39Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of Boulder. the Caribbean region; a plate-tectonic perspective: in 26 Ladd, J.W., Truchan, M., Talwani, M., Stoffa, P.L., Buhl, Dengo, G. & Case, J.E. (eds), The Geology of North P., Houtz, R., Mauffret, A. & Westbrook, G. 1984. America. Volume H. The Caribbean Region, 405-432. Seismic reflection profiles across the southern margin Geological Society of America, Boulder. of the Caribbean. Geological Society of America Mem- 40Pinet, B., Lajat, D., Le Quellec, P. & Bouysse, P. 1985. oir, 162, 153-159. Structure of the Aves Ridge and Grenada Basin from 27 MacDonald, K.C. & Holcombe, T.L. 1978. Inversion of multichannel seismic data: in Mascle, A. (ed.), Geody- magnetic anomalies and sea floor spreading in the namique des Caraibes, 53-64. Technip, Paris. Cayman Trough. Earth and Planetary Science Letters, 41Pszczolkowski, A. & Flores, R. 1986. Fases tectonicas del 40,407-414. Cretacico y del Paleogeno en Cuba occidental y central. 28 Mann, P. & Burke, K. 1984. Neotectonics of the Carib- Bulletin of the Polish Academy of Sciences, 34, 95-111. bean. Reviews of Geophysics and Space Physics, 22, 42Renne, P.R., Martinson, J.M., Hatten, C.W., Somin, M.L., 309-362. Onstott, T.C., Millan, G. & Linares, E. 1989.40Ar/39Ar 29 Mann, P., Draper, G. & Buike, K. 1985. Neotectonics of and U-Pb evidence for late Proterozoic (Grenville-age) a strike slip restraining bend system, Jamaica: in Bid - continental crust in north-central Cuba and regional die, K.T. & Christie-Blick, N. (eds), Strike Slip Defor- tectonic implications. Precambrian Research, 42,325- mation, Basin Formation, and Sedimentation. Society 341. of Economic Paleontologists and Mineralogists Spe- 43Rosencrantz, E. 1990. Structure and tectonics of the Yu - cial Publication, 37,211-226. catan Basin, Caribbean Sea as determined from seismic 30 Mann, P., Hempton, M., Bradley, D. & Burke, K. 1983. reflection studies. Tectonics, 9,1037-1059. Development of pull-apart basins. Journal of Geology, 44Rosencrantz, E. & Mann, P. 1991. SeaMARC II 91,529-554. mapping of transform faults in the Cayman Trough. 31Mann, P., Schubert, C. & Buike, K. 1990. Review of Geology, 19, 690-693.

Geologic Provinces of the Caribbean Region

45Rosencrantz, E. & Sclater, J.G. 1986. Depth and age of July 10th-l2th, 1985,270-280. the Cayman Trough. Earth and Planetary Science Let- 52Tomblin, J.F. 1975. The Lesser Antilles and Aves Ridge: ters, 79,133-144. in Nairn, A.E.M. & Stehli, F.G. (eds), The Ocean 46Ross, M.I. & Scotese, C. 1988. A hierarchial tectonic Basins and Margins. Volume 3. The Gulf of Mexico and model of the Gulf of Mexico and Caribbean region. the Caribbean, 467-500. Plenum, New York. Tectonophysics, 155,139-168. 53McCann, W.R. & Pennington, W.D. 1990. Seismicity, 47Sigurdsson, H., Sparks, R.S.J., Carey, S. & Huang, T.C. large earthquakes, and the margin of the Caribbean 1980. Volcanic sedimentation in the Lesser Antilles Plate: in Dengo, G. & Case, I.E. (eds), The Geology of arc. Journal of Geology, 88,523-540. North America. Volume H. The Caribbean Region, 48Sheridan, R.E, Muffins, H.T., Austin, J.A., Ball, M.M. & 291-306. Geological Society of America, Boulder. Ladd, J.W. 1988. Geology and geophysics of the Baha- 54Torrini, R., Jr. & Speed, R.C. 1989. Tectonic wedging in mas: in Sheridan, R.E. & Grow, J. A. (eds), The Geology the forearc basin-accretionary prism transition, Lesser of North America. Volume 1-2. The Atlantic Continental Antilles forearc. Journal ofGeophysical Research, 94, Margin, 329-364. Geological Society of America, 10549-10584. Boulder. 55Toussaint, J.F. & Restrepo, J. J. 1982. Magmatic evolution 49Simkin, T., Siebert, L., McClelland, L., Bridge, D., Ne - of the northwestern Andes of Colombia. Earth-Science whall, C. & Latter, J.H. 1981. Volcanoes of the World. Reviews, 18,205-213. Smithsonian Institution, Washington D.C. & Hutchin- 56Uchupi, E. 1975. The physiography of the Gulf of Mexico son Ross, Stroudsburg, viii+232 pp. and the Caribbean Sea: in Nairn, A.E.M. & Stehli, F.G. 50Somin, M.L. & Millan, G. 1977. Sobre laedad de las rocas (eds), The Ocean Basins and Margins. Volume 3. The metamorficas Cubanas. Academia de Ciencias de Gulf of Mexico and the Caribbean, 1-64. Plenum, New Cuba, Informe Cientifico-Tecnico, 2, 1-11. York. 51Speed, R.C. 1986. Cenozoic tectonics of the southeastern 57Wadge, G. & Wooden, J.L. 1982. Cenozoic alkaline vol- Caribbean and Trinidad: in Rodrigues, K. (ed.), Trans- canism in the northwestern Caribbean: tectonic setting actions of the First Geological Conference of the Geo- and Sr characteristics. Earth and Planetary Science logical Society of Trinidad and Tobago, Port-of Spain, Letters, 57,35-46.

CHAPTER 2

Evolution of the Gulf of Mexico and the Caribbean

JAMES L. PINDELL

Department of Earth Sciences, Fairchild Center, Dartmouth College, Hanover, New Hampshire 03755, U.S.A.

INTRODUCTION discussed:

FIXIST AND mobilist views of the evolution of Caribbean (A) The reconstructions of Pangea, including the restoration 63 region have both been proposed. Strictly fixist views are of syn-rift extension along passive margins during con difficult to entertain in light of the very well-documented tinental breakup; the bulk shape changes due to opening history of the Atlantic Ocean and the fairly accurate transcurrent and convergent faulting in northern South Pangean continental assemblies, both of which show that America during Andean orogenesis; and the removal of little or no Gulf of Mexico and Caribbean region existed Mesozoic -Cenozoic accreted arc terranes from northern between the larger continents during the Triassic, Jurassic South America for times prior to accretion. 14,48,49,65,70,81 and early Cretaceous . Mobilist views all ac cept (B) Atlantic opening kinematics and implications for North- significant amounts of eastward Caribbean migration South America motion. relative to the Americas, but are split between models which (C) Mesozoic kinematic significance of the Equatorial At- generate the lithosphere of the Caribbean Plate between the lantic reconstruction. 6,29,42,48,82 Americas and models which generate that litho- (D) The opening history of the Gulf of Mexico. 27,30,69,72 67 sphere in the Pacific . Pindell listed cogent arguments (E) The eastward migration of the Caribbean Plate from the for the plate's Pacific origin, but definitive proof will only Pacific, independent of Cayman Trough magnetics, by come when the deep interior, and not just the rims, of the tracing the timing of overthrusting of circum-Caribbean Caribbean Plate is shown to be pre-Aptian in origin, as foredeep basins by Caribbean terranes. plate reconstructions dictate that the Caribbean Plate could (F) The occurrence of magmatic arcs, and their periods and not have fitted between the America's until well after that polarities of associated subduction. time (see below). However, I note that the recent identifica- (G) Arc-continent collision suture zones marking the sites tion of a Jurassic boreal or austral radiolarian fauna in 62 of former oceans/basins, and their timing and vergence Hispaniola, Puerto Rico, and La Desirade also attests to of closure. the Pacific origin, as the Proto-Caribbean Seaway of Pin- 65 (H) The opening histories of the Cayman Trough, Grenada dell developed essentially within the Jurassic palaeo- Basin and Yucatan Basin. equatorial zone. (I) Plate boundary zone development in the northern and Important elements of (1) the methodology required for southern Caribbean. regional analysis, and (2) the actual history of the Caribbean inter-plate realm (Fig. 2.1), were outlined in progressively The present paper is a summary of an evolved plate more detail by, among others, Ladd , Pindell and Dewey , 55 65 14 tectonic model for the Gulf of Mexico and Caribbean region Mann and Burke , Pindell , Buffler and Sawyer , Dewev by Pindell et al76. The summary utilizes many of the above and Pindell26, Klitgord and Schouten48, Pindell et al.70, 79 16 81 elements without repeating them, except where appropriate. Rosencrantz et al. , Burke , Rowley and Pindell , Pindell The summary of the model is supported with a limited and Barrett69 (and other papers in Dengo and Case22), 78 68 76 amount of local detail to point out how various aspects of Rosencrantz , Pindell and Pindell et al . These papers local geology are explained by the model. Geographic areas

14 JAMES L. PINDELL

Figure 2.2. Plate kinematic history between North and South America, Triassic to present (after Pindell et al.70) Vectors denote flowlines travelled by points on South America relative to North Americ a. where commonly accepted data are discordant to the model tonic and palaeogeographic evolution. can be considered as candidates for future research, to either reassess those data or to modify the general model. Ulti- mately, constraints derived by the deductive (tectonic mod- PLATE KINEMATIC CONSTRAINTS els) and inductive (interpretation of field data) approaches of assessment should merge and agree in the future to Atlantic Ocean Magnetics produce a common interpretation of the region's plate tec- Interpretations of the geologic evolution of the Carib-

Figure 2.1. (opposite) General location and basin map of the Caribbean region and sites mentioned in text (after Pindell68). Key to Proto-Caribbean/Caribbean episutural foredeep basins: 1, Sepur foredeep basin (and Chiapas foldbelt), Guatemala and Belize; 2, Cuba-Bahamas foredeep basin; 3, Maracaibo foredeep basin, Colombia and Venezuela; 4, Eastern Venezuelan/Trinidad foredeep basin. Key to rift and pull-apart basins: 5, Yucatan basins (lithospheric rift); 6, Grenada basin (lithospheric rift); 7, Eastern Belize margin basins; 8, Cayman Trough basin (lithospheric rift); 9, Nicaraguan Rise basins (probably upper crustal grabens); 10, Falcon basin (lithospheric? rift); 11, Sambu Basin, Panama; 12, South America borderlands basins (Baja Guajira, Triste, Cariaco, Bonaire, La Vela, Carupano, Leeward Antilles inter-island basins) (probably upper crustal grabens, but Bonaire basin probably had Oligocene lithospheric extension like Falcon). Key to accretionary prism/forearc basins: 13, Barbados Ridge and Tobago Trough basins; 14, South Carib-bean/Panama/Sinu-San Jacinto foldbelts and Lower Magdalena Basin; 15, San Juan-Azua-San Pedro- Enriquillo Basin, Dominican Republic; 16, Nicoya complex, Costa Rica. Key to arc-flank and other basins: 17, Limon basin, Costa Rica; 18, North Puerto Rican basin; 19, South Cuban shelf; 20, Cibao basin; 21, Cesar basin, Colombia; 22, Saba Bank platform. Oil fields shown as blobs.

15 The Gulf of Mexico and the Caribbean

NOAM - SOAM SPREADING HISTORY

Figure 2.3. Rates of relative motion for North-South American displacement, from data shown in Figure 2.2 and discussed in text. The two rates after the Eocene for Maracaibo and Trinidadian portions of northern South America are due to clockwise rotation of that plate relative to North America for that time. bean must be set in the framework of the former relative Ocean closed, is, therefore, the natural starting point for positions and motions of the encompassing North and South models of evolution of the Caribbean region. During the late American Plates. This is made especially poignant by Triassic -Jurassic rifting and subsequent drifting, the history Permo-Triassic reconstructions of North America, South of seafloor spreading in the oceans defines the size and shape America and Africa which show that the Caribbean region at any instant of the Caribbean inter-plate realm. Three did not exist at that time. The Caribbean must therefore have segments of the Atlantic, the central north, the south and the evolved as a co-development of the dispersal of these con- equatorial, are important for determination of the Caribbean tinents. Relative plate motion studies, as measured by mag- kinematic framewoik. Marine magnetics and fracture zones netic anomalies and fracture zone traces, are accurate for of the central north Atlantic between the U.S.A. and north- each palaeoposition to a few tens of kilometres. In contrast, west Africa/South America define the history of separation attempts to define the Caribbean framework by assessments of North America from Gondwana (two-plate system only) of the latitudinal component of motion between North and from the late Triassic to the Aptian. It was during this stage South America, measured by onshore determinations of that Yucatan rotated from its Triassic, Pangean location to palaeo-inclination through time, are less accurate by at least its present position to form the Gulf of Mexico by or just an order of magnitude. The kinematic framework in which after anomaly M-16 (Berriasian) time, which is the first time 65 Caribbean evolution took place is shown in Figures 2.2 and at which overlap with South America can be avoided . 2.370. The poles determined for the construction of Figure Also, wfthin the Neocomian, spreading began in the south 2.2 fall within error estimates of more recent pole determi- Atlantic, but significant motion through the equatorial At- nations85. lantic appears to have been delayed until the Aptian. Prior An accurate reconstruction of pre-Mesozoic continen- to this the early south Atlantic motions were manifested tal fragments for Permo-Triassic time, with the Atlantic northward into the Central African rift system rather than

16 JAMES L.PINDELL

the equatorial Atlantic 70'72. Therefore, early south Atlantic sphere of the arm of the Atlantic called the Proto-Caribbean motions do not appear to have significantly affected Carib- Seaway by Pindell65; or it was generated in the Pacific bean kinematic history. (Farallon Plate lithosphere?), such that Proto-Caribbean Unfortunately, the Cretaceous magnetic quiet period crust which was already formed by the separation of the prevents resolution of detailed kinetics for Aptian to San- Americas was then subducted beneath the Upper Cretaceous tonian times. Opening poles for the central north and south to Cenozoic arc systems of the Caribbean Plate during the Atlantic Ocean show that by the early Campanian and until westward drift to the Americas from Africa, thus producing a the Eocene (anomaly 34 to anomaly 21, that is, 84-49 Ma), reactive east-west migration history of the Caribbean Plate little or no motion was occurring between North and South between the Americas. In either case, westward drift of the America, and it is likely that no significant plate boundary Americas across the mantle was mainly responsible for existed between them for that interval70. An important un- east-west Caribbean relative motion. In the case of a Pacific certainty is the exact time at which seafloor spreading actu- origin, northerly and southerly extensions of Farallon litho- ally ceased in the Proto-Caribbean. It certainly had ceased sphere were probably subducted beneath the North and by anomaly 34 time (as shown by the dashed portion of the South American Cordilleras, respectively, thereby produc- curve in Fig. 2.3). I suggest that the exceedingly rapid ing the condition of tectonic rafting of Caribbean lithosphere Albian-Cenomanian transgression of cratonic areas (for ex- into the Proto-Caribbean Seaway between the Americas. ample, in Venezuela37) and drowning of carbonate plat- The primary difference of these two interpretations lies forms was aided by loss of in-plane stress as a result of the in the magnitude of the relative east-west migration of death of the ridge71,76. This is supported by the occurrence Caribbean and American lithospheres. Therefore, this dif- of oceanic crust west of anomaly 34 immediately east of the ference suggests different locations for much of the Carib- Lesser Antilles in the western Atlantic, which has the same bean region's Jurassic and Cretaceous magmatism, fabric orientation as that to the east of anomaly 3489, imply- sediment deposition, deformation and metamorphism. In ing that any adjustments in orientation associated with the models for a Pacific origin, early Caribbean stratigraphies death of the ridge (reorganization to a two-plate system and tectonism must have developed in the Pacific prior to again) had already taken place well before anomaly 34 time the relative eastward migration, and are thus essentially part (late Albian?). If this is the case, then the period over which of 'Cordilleran' evolution. In Proto-Caribbean models of the African and North and South American plates behaved Caribbean Plate origin, such developments are strictly 'Car- as a three-plate system was limited to the Aptian-Albian ibbean', and should have involved the Yucatan, Bahamian, interval. This indicates that the opening history of both the and northern South American cratonic margins. central north and south Atlantic were essentially co-polar Pindell67 outlined several independent arguments fa- (that is, one greater American Plate) from the ?Albian to the vouring a Pacific origin. Briefly, these are (see Fig. 2.4): Eocene, although some minor degree of wrenching probably occurred at Atlantic fracture zones, possibly along those 1. Eastern Caribbean Volcanism. The Aves Ridge and extending to the Bahamas. Lesser Antilles volcanic arc complexes (Fig. 2.1) collec Since the middle Eocene, very slow north-south con- tively possess an Upper Cretaceous (about 90 Ma) to Recent vergence (dextrally oblique relative to pre-existing fracture record of intermediate arc magmatism. Polarity of subduc- zones) occurred, with the magnitude of convergence in- tion for the Eocene-Recent Lesser Antilles arc has been creasing westward away from the North America-South eastward facing with westward dipping subduction, as was America pole of rotation to the east of the Lesser Antilles probably the case for the Cretaceous-Eocene Aves Ridge arc (Figs 2.2,2.3). It is not yet known whether this convergence as suggested by its slightly convex shape and absence of an was accommodated only by descent, and maintenance of accretionary prism along it west flank. Assuming a 90 north-south slab distance, of Atlantic lithosphere into the million year period of west-dipping subduction of Atlantic Benioff zone beneath the eastward migrating Caribbean crust beneath the eastern Caribbean even at only slow con Plate, or if lithospheric shortening at an overthrust zone east vergence rates suggests a minimum relative plate movement of the migrating Caribbean Plate was also required, possibly of circa 1000 km. 75 along northern South America . 2. Cayman Trough, seismic tomography, and northern Caribbean strike-slip basins. Assessments of the develop Origin of the Caribbean Plate ment of the Cayman Trough79,93, the Tabera and northern Within the above framework, there are two possibilities San Juan Basins in Hispaniola28,56, the 'Eocene Belt' of for the origin of the lithosphere of the Caribbean Plate. Puerto Rico36, and the Cibao Basin and north coastal area Either it was generated by seafloor spreading between Yu- of Hispaniola34,73 (Fig.2.4) indicate late Eocene to Recent, catan and South America and, therefore, represents litho- east-west, sinistral strike-slip motion between the Caribbean

17 The Gulf of Mexico and the Caribbean

Figure 2.4. Summary of arguments for a Pacific origin of the Caribbean Plate, after Pindell67. Light lines denote approximate present-day general boundary configuration. Heavy lines denote approximate boundary (suture zones that formed during Caribbean evolution) between the allochthonous Caribbean arc/oceanic rocks, and the autochthonous and para-autochthonous Proto-Caribbean passive margin/foredeep basin rocks. Eastern Caribbean volcanism in Aves Ridge and Lesser Antilles (argument 1) shown with Vs. Post-Middle Eocene strike-slip basins (argument 2) shown as A (Cayman Trough), B (San Juan Basin), C (Tabera Basin), D (Eocene Belt). General stratigraphic differences of Caribbean and Proto-Caribbean rock suites (argument 3) shown in inset. Aptian position of South American outline and shelf relative to North America, showing small separation between the Americas at that time (argument 4) labeled 'Aptian So Am'. Sequential migrated positions of the Chortis block and the trench-trench-transformed (TIT) triple junction relative to Mexico (argument 5) shown by outlines southwest of Mexico: mainly Paleogene and Neogene/Quaternary arcs shown in ‘v’s. Division between pre-Campanian Pacific and Proto-Caribbean faunal realms (argument 6) nearly matches suture zones outlined in heavy lines. Foredeep basins that developed ahead of the relatively eastwardly migrating Caribbean Plate shown by dotted areas, where SF=Sepur Foredeep (late Cretaceous); CF=Cuban Foredeep (early Paleogene); MF= Maracaibo Foredeep (Eocene); EVF=Eastern Venezuelan or Maturin Foredeep (Miocene). and North American plates (northern PBZ). Offset of about 3. Caribbean vs. Proto-Caribbean stratigraphic suites. 100 km since the Eocene is indicated from the length of the Cretaceous portions of the stratigraphies of two distinct deep oceanic portion of the Cayman Trough, and from suites of rock in the Caribbean region (Figs 2.4, 2.5) are reconstructions of Cuba, Hispaniola, Puerto Rico and the genetically incompatible as presently juxtaposed across cir- Aves Ridge arc fragments into a single, pre-middle Eocene cum-Caribbean ophiolite belts interpreted to be suture zones Greater Antilles arc69. Seismicity61 and seismic tomogra- (Fig.2.4). The 'Caribbean suite' occurs 'Caribbeanward' of phy39 show a distinct west-dipping Atlantic Benioff zone the suture zones, whereas the 'Proto-Caribbean suite' occurs extending at least 1,200 km beneath the eastern Caribbean, 'Americanward' of the sutures. The Caribbean suite's Cre- suggesting a similar minimum magnitude of displacement taceous, tuff-dominated stratigraphy differs dramatically as the Cayman Trough. If this much motion has occurred from the Proto-Caribbean suite's coeval Cretaceous, non- since the Eocene, a much larger value for the total relative volcanogenic, passive shelf sediments. Spatial separation motion must have occurred as indicated by the late Creta- during deposition of these distinct suites of rock appears ceous period of arc activity of the Aves Ridge. necessary for the Proto-Caribbean to contain no record of

18 JAMES L. PINDELL

Caribbean volcanism. As this difference extends from Campanian time44, suggesting spatial separation of shal- around the Caribbean Sea to the Santa Marta massif of low-water organisms prior to that time. The areas of occur- Colombia and to Chiapas, Mexico, it is unlikely that the rence for the two realms closely match the areas of the Caribbean arcs were situated any farther east than these Caribbean and Proto-Caribbean stratigraphic suites. The locations for most of the Cretaceous. Campanian initiation of faunal merging of the realms relates 4. Pre-Aptian geometrical incompatibility, Caribbean to the onset of tectonic juxtaposition of the shelfal areas they Plate and Proto-Caribbean Seaway. Numerous faunal and occupied, presumably during relative eastward migration of isotopic ages from the basements of most Caribbean arcs are the Caribbean Plate between the Americas. Further, pre-Aptian, and seismic sections of the Colombian and Montgomery et al62 identified cold water forms of Upper Venezuelan Basin90, particularly the basement continuity Jurassic radiolarians in the Puerto Plata Basement Complex from the Jurassic rocks of Costa Rica to the Colombian of Hispaniola73, the Bermeja complex of Puerto Rico59, and Basin, suggest that the crust of the internal Caribbean plate on La Desirade, which can only be explained by a Pacific, is pre-Aptian (probably Jurassic) as well. However, plate non-Tethyan, original for the basements of those localities. separation between North and South America was insuffi The above arguments collectively indicate that the crust cient (Fig. 2.4) to house a pre-Aptian Caribbean Plate until of the Caribbean Plate and the Chortis block was situated the late Cretaceous, probably the Albian. Thus, the Carib west of the Cretaceous shelf sections of Yucatan, the Baha- bean Plate must have formed outside the present Caribbean mas, and northern South America prior to the later Creta- area, much farther west than the 1100 km of displacement ceous (Campanian). Seafloor spreading between North and indicated by the Cayman Trough and seismic tomography, South America had ceased, probably in the Albian, leaving and its migration history must have began well before the a Proto-Caribbean oceanic arm of the Atlantic to subside Eocene, probably in the early late Cretaceous as suggested thermally in the absence of plate boundaries. In addition to by Aves Ridge subduction-related volcanism and by the understanding regional evolution, acknowledgement of the Albian to early Tertiary ‘Antillean phase’ of magmatism in existence of this Proto-Caribbean Seaway, with the Carib- the Greater Antilles. bean Plate situated to the west, is critical to the hydrocarbon 5. Truncation and uplift of the southwestern margin of story of the circum-Caribbean region because it was along Mexico. Structural trends and the Paleogene arc of south this seaway's margins that the region's best source rocks west Mexico have been truncated46 either by subduction were deposited, from Albian to Campanian time68. erosion or by strike-slip removal of arc and forearc areas. In light of these arguments, the concept of the present- Paleogene arc rocks in Mexico are largely restricted to the day Caribbean lithosphere representing a piece of the Proto- Sierra Madre Occidental; the Trans-Mexican Volcanic Belt Caribbean Seaway is difficult to entertain. The implication arc is mainly Neogene in age, with volcanism commencing is that the entire Caribbean Plate is allochthonous relative to earlier in the west than in the east, implying an eastward the Americas and has migrated from well over 200 km to migration of arc inception (Fig. 2.4). In the Chortis Block the west, and that (what are now) adjacent portions of the of Central America, evidence for Paleogene arc activity is Mesozoic Caribbean and America (Proto-Caribbean) strati- abundant, with volcanism extending back into the Creta graphic suites should not be correlated due to the large ceous. The Paleogene arc sequences were probably continu spatial separation during original deposition. Deep drilling ous from western Mexico into Chortis, prior to eastward within the interior of the Caribbean Plate would likely return strike-slip offset of Chortis to its present position and pro an undeformed stratigraphic sampling of the eastern Pa- gressive development of volcanism in the Trans -Mexican cific's Upper Jurassic(?) and Lower Cretaceous section, Volcanic Belt93. In addition, Precambrian rocks of the which elsewhere is only poorly preserved. southwest margin of Mexico46 yield cooling ages that indi- Merging the relative motions between North and South cate uplift and erosion since Oligocene time, younging America (Fig.2.2) with a Jurassic or early Cretaceous eastwards21. This probably occurred as a function of an Pacific origin of the Caribbean lithosphere implies a very intra-continental, strike-slip fault zone (between Mexico simple history of Caribbean evolution that can be described and Chortis) progressively becoming a Neogene subduction generally by two phases. The first phase was Triassic to mid margin as Chortis migrated to the east, with associated uplift Cretaceous, northwest-southeast relative separation of of the hanging wall (Mexico). Thus, it appears that Chortis North and South America, and the opening of the Proto-Car- has migrated with the Caribbean crust during Cenozoic time ibbean Atlantic-type seaway bounded by passive margins frpm a more westerly position72,93. along the Bahamas, eastern Yucatan and northern South 6. Faunal provinciality: Pacific versus Proto-Caribbean America. The second phase involved the Albian to Recent Realm. Two differing Cretaceous faunal realms exist across subduction of that Proto-Caribbean lithosphere beneath arcs the Mexican-Caribbean region that remained distinct until along the eastern Caribbean border, during westward drift

19 The Gulf of Mexico and the Caribbean

Figure 2.5. Generalized stratigraphic columns for portions of the Proto-Caribbean passive margins versus the central portion of the Caribbean Plate (after Pindell8).

of the America's from Africa. Eastwardly progressive de- assessing total Tertiary strike-slip displacements, because struction of the Proto-Caribbean passive margins by the they were not in place at the onset of the strike-slip disloca- relative motion of the Caribbean Plate would be recorded by tions. In the simple two-phase model, the emplacement of foredeep basin development above the pre-existing Proto- allochthons occurs as a direct consequence of mainly Terti- Caribbean shelf sections. Figure 2.4 shows four large basins ary, Caribbean-South American relative motion. Therefore, formed by this process whose eastwardly-younging ages of most of the Caribbean-South American transcurrent offset foredeep loading are: Sepur, Guatemala (Campanian-Maas- occurred along the basal thrusts of the allochthons. The high trichtian); Cuban (early Paleogene); northern Maracaibo angle strike-slip faults within the orogen, which are secon- (Eocene); and Eastern Venezuelan (Miocene). dary in magnitude of offset, thus record only a minor portion The often-cited Cretaceous orogenesis along northern (less than 150 km) of the total relative motion which is in South America9,10,19,57 does not fit the simple scenario of excess of 1,000 km in the vicinity of Guajira Peninsula. Caribbean-South American interaction26'66 implied by the However, it is apparent that as one heads east, the actual total above kinematics, and this discrepancy has been the subject offset between South American and Caribbean terranes of much recent study. No Cretaceous orogenesis affected becomes progressively less, because the Caribbean-South Trinidad1-3. Similarly, passive margin conditions prevailed America plate boundary did not exist until progressively across northern South America until the Tertiary 74,75. An younger times toward the east. implication of these concepts is that all rocks in northern South America affected by metamorphism in the Cretaceous CARIBBEAN EVOLUTION: PHASE 1; NORTH are Caribbean-derived and allochthonous. Flysch units con- AND SOUTH AMERICA DRIFT STAGE taining South American shelf debris, once believed to be Cretaceous in age on the basis of clasts, are now known to The following discussion on regional Caribbean evolution have Tertiary matrices (Gaitapata and Paracotos Forma- is adapted from a full synthesis by Pindell et al.76. The first tions), in keeping with the simple, two-phase tectonic phase of evolution (Triassic to Albian) was primarily asso- model. A second important implication is that the belts of ciated with the development of the Proto-Caribbean Sear Cretaceous metamorphic rocks are not viable markers for way. The second phase (Albian to Recent) involved the

20 JAMES L.PINDELL progressive consumption of Proto-Caribbean crust beneath Ma). The crust to the north of the Blake Plateau that was Caribbean arcs during westward drift of the Americas across the created by seafloor spreading prior to formation of the mantle, leading to the present plate configuration. BSMA produced the central north Atlantic's asymmetry with respect to the present ridge axis. An early ridge appar- Late Triassic-Jurassic ently was abandoned as the spreading centre jumped to the The western Pangean reconstruction of Figure 2.6a is site of the BSMA86,92. modified after Pindell65 and Rowley and Pindell81. The The northeastern Gulf of Mexico, Florida, the western total closure pole for Yucatan/North America relative to Florida Shelf, the Blake Plateau and the western Bahamas is 14,47,65 North America is latitude 28.4, longitude -82.1, angle -47.781 a complex region of continental blocks . Sinistral and lies in northern Florida. The Atlantic closure poles are motion along the Jay Fault began such that continental hosts after Pindell et al.70. The Chiapas Massif is not included with the (Sabine, Monroe, Wiggins, Middle Grounds or Florida El- Yucatan Block. Andean deformations in northwest South bow, and Tampa or Sarasota Arches) became separated by America have been restored in a similar (but more rigorous) rifts along the northeastern Gulf Coast margin. Extension way to that of Dewey and Pindell26. Mexican terranes have was far greater south of the Jay than it was to the north, been displaced by the minimum amount necessary to avoid requiring a sinistral component of displacement in addition to overlap with South America. Chortis is not included along normal offset at essentially a transfer zone. Total offset, western Colombia as it is in some reconstructions because which continued into the Jurassic, increased southwards the Central Cordillera of Colombia possessed a Triassic - Ju- relative to North America (differential shear) and perhaps rassic arc axis and must have been adjacent to the Pacific was as much as 100 to 150 km between the Sarasota Arch lithosphere. and the Ouachitas, the latter of which serves as a fixed, North In late Triassic to early Jurassic time, North America America reference frame. Thus, the North Louisiana and began to rift from Gondwana along widespread, poorly-de- Mississippi Salt Basins (although no salt was yet deposited), fined zones of intracontinental block-faulting, redbed depo- the Northeast Gulf Basin and the Florida Elbow Basin sition and dyke emplacement (Fig. 2.6a, b). In the southern between the aforementioned highs came into existence. 65 U.S.A. , north-south extension77 produced an extension gra- Pindell noted that there is (a) sufficient overlap between ben system filled with Eagle Mills redbeds. Along the the present day limit of continental crust in the Bahamas and eastern U.S.A., redbeds, dykes and sills of the Newark that in the Guinea Plateau of western Africa during Pangean Group and its equivalents were deposited and emplaced in a closure, and (b) an unfilled gap in the eastern Gulf of Mexico belt of grabens paralleling the coast from the Piedmont out to during closure, to warrant the suggestion that a fault zone in the continental shelf. The Eagle Mills and Newark systems addition to the Jay exists between the Florida Elbow and are lithologically and tectonically equivalent. These systems Sarasota Arches (Florida Elbow Basin) along which crust of relate to hanging wall collapse of pre-existing thrust faults of south Florida and the Bahamas migrated eastwards during the Alleghanian Orogen, as the latter went into extension. Gulf opening. Although the Guinea overlap could be ex- Judging from subsidence history, in which post-rift thermal re- plained also by magmatic addition during crustal stretching in equilibration is lacking, these basins are upper crustal the South Florida Volcanic Belt, this still does not satisfy detachments rather than true lithospheric rifts. The true the Gulf gap problem, suggesting that one or more faults lithospheric rifts developed farther south and east, cross Florida to allow the marginal offset between the Ba- 15 respectively, during the Jurassic. hamas and the Blake Plateau. Buffler and Thomas showed Drift between Africa and North America was underway in faulted basement at the west flank of the Florida Elbow the middle Jurassic (Fig. 2.6b). The Punta Alegre and Basin which could coincide with the Florida Elbow Fault 65 Exuma Sound(?) salt of the Bahamas might correlate with zone of Pindell , but the strike of the faults is not clear. In the portions of the Louann and Campeche in the Gulf of any case, it is noted that assessments of relative motion Mexico, but these units are probably older and appear to between North America and the Yucatan Block based on relate to the opening of the central north Atlantic rather than structural or tectonic style cannot be made in the eastern to the Gulf of Mexico. Such is the case with the Senegal Gulf during the period over which the blocks south of the Basin and Guinea Plateau salt deposits43 which opposed the Jay Fault, including the Sabine and Wiggins Arches, were Bahamas at that time. The Blake Spur Magnetic Anomaly moving, because the blocks were independent, intervening 65 (BSMA; Fig. 2.6b) can be symmetrically correlated about the terranes with their own relative motion . mid-Atlantic Ridge with the western margin of Africa, but To the west, the Yucatan Block began to migrate south- internal stretching within the Blake Plateau accounts for wards along the eastern flank of the Tamaulipas Arch, divergence between North America and Africa until the time producing the arcuate shear zone along eastern Mexico when the BSMA formed to the north (approximately 170 (Tamaulipas-Golden Lane-Chiapas transform fault of Pin-

21 The Gulf of Mexico and the Caribbean

65 dell , not a rifted margin) which helps define the North Legend for Caribbean Evolutionary Maps America-Yucatan pole of rotation in Florida (Fig. 2.6b, c). This shear zone truncates any faults entering the Gulf from Mexico, including hypothetical extensions of the Mojave- LEGEND FOR CARIBBEAN Sonora trace, and therefore must have remained active EVOLUTIONARY MAPS longer. The absence of a major marginal offset (about 700 km6,48) on basement isopach maps of eastern Mexico15 precludes the possibility that a Mojave-Sonora fault zone entered the Gulf of Mexico during rifting, and renders unlikely all tectonic models which open the Gulf by moving Yucatan along Atlantic flowlines. The question is, then, how did the terranes of Mexico get into the South American overlap position after rifting? One possibility is that strike slip systems such as the Mo- jave-Sonora and the Nacimiento may have been active during the Jurassic toward the northwest, but, rather than entering the Gulf of Mexico to the east, they instead trended more southwards to the west of the Tamaulipas Arch. Un- fortunately, this area has been overthrust by the Sierra Madre Oriental so that this hypothesis is not easily tested. However, the suggestion would provide a logical mechanism for the production of the deeper water depocentre of the section now comprising the Sierra Madre. I suggest that the southward trace of the fault system(s) became transtensional in central Mexico, creating a Mexican backarc trough. This trough may have formed in order to maintain the subduction zone outboard of Mexico. In the southwest U.S.A., sinistrally transpressive Nevadan orogenesis oc - cured at this time due to migration of North America from Gondwana, but the margin presumably became progres - sively more oblique toward the south along Mexico (Fig. deeper water deposition across central Mexico, the depocen- 2.6c). Backarc extens ion may have been driven by subduc- tre for the Sierra Madre section; and fit very well into our tion zone rollback at the trench, and a significant component perception of plate kinematics of the region at that time. The in the backarc basin is predicted in order to help maintain proposed basin probably began its formation in the Cal- trench-normal subduction at the trench. In this way, the lovian, concurrent with the earliest known marine spills into terranes of Mexico could have migrated southwards, but in the Gulf of Mexico in the area (Tepexic Formation? in a more extensional direction relative to the Atlantic flowli- southern Mexico). Depending on the degree and variability nes. Later, during late Cretaceous -Eocene shortening in the of the obliquity of opening, the basin could have had vari- Sierra Madre thrustbelt, which was directed essentially to- able water depths, with limited basaltic intrusion or produc- ward the eastnortheast, the terranes encroached upon the tion of oceanic basement. Gulf of Mexico. The net travel path of southward transten- In northern South America, rifting, igneous intrusion sion followed by eastnortheast thrusting (two sides of a and redbed deposition occurred over Triassic -Jurassic triangle) may have produced an apparent transcurrent offset times. Rifts can be divided between those east of Maracaibo (third side of the triangle) along the hypothetical southeast- (Takatu, Espino, Uribante/Tachira, Cocinas?), which are ward extension of the Mojave-Spnora trace. This suggested due to continental breakup from Yucatan/Florida-Bahamas, history would: explain basement offsets in northwest Mex- and those west of Maracaibo (Machiques, Cocinas?, Bo- ico and southwest U.S.A. along the Mojave-Sonora and gota/Cocuy, Santiago, Payande), which are Andean backarc other faults6; provide a mechanism for Mexican terranes to basins with arc-related volcanism, as well as rift-related migrate into the South American overlap position in volcanism, in or adjacent to them. The Machiques, Pangean reconstructions; account for Sierra Madre shorten- Uribante/Tachira and Bogota/Cocuy rifts would become the ing; avoid predictions of a large and missing marginal offset sites of Andean uplift during Neogene times (Perija, Merida, along the eastern Mexican margin; suggest a mechanism for Eastern Cordillera, respectively) '41. The continental mar -

22 JAMES L. PINDELL

NORTH AMERICA

Figure 2.6a. Palaeogeography, late Triassic -early Jurassic.

Figure 2.6b. Palaeogeography, Bathonian.

23 The Gulf of Mexico and the Caribbean

gin to the north also formed at this time, as late and possibly of motion must have approximately summed to match the middle Jurassic marine, passive margin sediments were opening rate and azimuth of the central Atlantic Ocean. By deposited from Guajira to Trinidad3,37.The platform areas Oxfordian times, rifting appears to have ceased between the between these rifts behaved as basement highs during sub- blocks south of the Jay Fault, such that Gulf of Mexico sequent late Jurassic and early Cretaceous thermal subsi- marine magnetics and structural trends west, but not south- dence. The backarc marginal trough predicted for Mexico east, of the Florida Elbow Arch probably define Yucatan- (see above) is also predicted for offshore Colombia Triassic North America motion. Hall et al38 suggested a pole for this to Jurassic arc magmatism ceased in this area by the end of time based on magnetics that is farther south in Florida than the Jurassic and all of autochthonous Colombia became a my total closure pole defined earlier, and it may be that 74 part of the northern South American passive margin . The crustal stretching during the rift stage in the Gulf Coast was plate vector circuit for North America, South America and northwest-southeast directed, followed by more north-south Cordilleran Mexico can be satisfied by a single RRR triple rotational seafloor spreading. If so, the two stages of opening junction within the backarc basinal area, connecting the may sum match the total closure pole. Given, among Mexican, Colombia and Proto-Caribbean zones of displace- other things, the slight bend to the northwest in basement ment (Fig. 2.6c, d). This allowed continuity of the Cordille- structure contours in the Burgos Basin area at the north end ran arc systems in an outboard position away from passive of the Tamaulipas Arch15, this two stage model appears to margin elements of Colombia and eastern Mexico: the back be justified58. To the north of the migrating junction be- arc systems would collapse during the late Cretaceous tween the Tamaulipas-Golden Lane-Chiapas fault zone Sevier-Peruvian orogenesis after Aptian-Albian onset of (TGLC) and the central Gulf ridge system, the TGLC westward drift of South America from Africa and plate re- evolved as a fracture zone separating regions of differential organization in the eastern Pacific. subsidence. The abrupt topographic low, or freeboard, east By the middle Oxfordian, intra-continental extension in of TGLC has received enormous volumes of sediment the Gulf of Mexico had reached a point (constrained South through time (for example, Burgos Basin east of Monterrey, America divergence rate) where it had opened enough to Mexico46). As strike-slip motion ceased along the Tamaulipas accommodate the entire extent of the Louann and Campeche Arch after passage of the ridge system, it thermally evaporite basin, but salt deposition may have been in the subsided and eventually was onlapped by uppermost Jurassic Callovian or earlier. Oxfordian salt deposition is supported and Cretaceous carbonates. In Chiapas, deposits of the by seismic studies which show no erosion of the salt below Todos Santos 'rift facies' are younger (late Jurassic -early the Oxfordian Norphlet and Smackover Formations, sug- Cretaceous) than those in the northern Gulf. This is prob- gesting very little time between the deposition of the two ably because shear along the active TGLC continued longer in units . In most places, the salt appears to cover or mark the the south, well into the thermal subsidence stage of the breakup unconformity around the Gulf: elevation and expo- north. sure of the surface across the Gulf Basin prior to salt The whereabouts of the Chortis Block relative to North deposition was probably controlled by the ratio of crustal to 24 America during the Jurassic and early Cretaceous is un- lithospheric thickness until a critical value was reached known. However, by Aptian times the carbonate units of during continued extension. The salt often marks the onset southern Mexico and Chortis became very similar, and it of thermal subsidence which led to establishment of open could be argued that the Cretaceous magmatic rocks of marine conditions around most of the Gulf, but isolated Chortis formed a southerly extension of the Cretaceous pockets may have been well below sea level during the Mexican arc as well. For simplicity, I do not show the Oxfordian advance of the seas, resulting in rapidly deepen- Chortis Block in the reconstructions until the Valanginian. ing conditions. Two likely entrances for marine waters into To the east, the juvenile Proto-Caribbean continued to the Gulf during the Callovian-Oxfordian times are between open, fanlike, between Yucatan and Venezuela. The mid- Florida and Yucatan, as DSDP leg 77 documented Jurassic 83 ocean ridge in this basin must have been joined in some way extension and marine sedimentary rocks , and the southern with the plate boundary separating the Florida/Bahamas and Mexican Isthmus, where Callovian marine rocks of the Yucatan Blocks, which in turn have been connected to the Tepexic Formation occur. Central North Atlantic system by a long transform along the During the entire late Jurassic, seafloor spreading was south side of the Bahamas. A single triple junction is por- occurring in the Gulf of Mexico where it separated the salt trayed connecting these plate boundaries in the northern basin into halves, and also in the Proto-Caribbean Sea (Fig. Proto-Caribbean, for simplicity, but late Cretaceous to Pa- 2.6c). The crust of southern Florida/Bahamas may have leogene subduction of this crust has eliminated direct evi- been mobile relative to North America as well, by faulting dence for this proposition. The crust east of Yucatan and along the Florida Elbow Fault Zone. The three components south of the Bahamas may have had a rough bathymetric

24 JAMES L. PINDELL

Figure 2.6c. Palaeogeography, late Oxfordian.

Figure 2.6d. Palaeogeography, Valanginian.

25 The Gulf of Mexico and the Caribbean character (stretched continental blocks or raised oceanic gesting that the trenches of whatever arc systems existed fracture zones?), such that a complex pattern of carbonate west of the Proto-Caribbean area did not intersect the South highs and lows developed into the Cretaceous. Sediments America Andes until southern Ecuador. I speculate that an originally deposited in the northern Proto-Caribbean depo- evolving and lengthening intra-oceanic arc system from centre is now represented in the Cuban thrust belt north of Chortis/Mexico to southern Ecuador would form the roots and underneath the Cuban ophiolite/arc belt. Siliciclastic to the Greater Antilles and Aves Ridge arcs of the Caribbean sands in those sections were probably derived laterally from and the Amaime-Chaucha terrane of Colombia/Ecuador, Yucatan rather than the Bahamas. Hence, sedimentary evi- although at this point the arc was westward facing. Subduc- dence in Cuba for the Cuba-Bahamas collision appears older tion of Pacific (Farallon?) lithosphere at this speculative arc (late Cretaceous) than it actually was (Paleogene). configuration may have been accompanied by accretion of The opening of the Gulf of Mexico was achieved prob- Jurassic ophiolitic fragments and cold water cherts now ably within the Berriasian, when sufficient room existed found in Dominican Republic, Puerto Rico, and La 62 between North and South America to avoid overlap between Desirade . The relative velocity triangles of Figure 2.6d Yucatan and South America (anomaly M-16). At that time, describe the relative plate motions at this time. the Yucatan block became part of the North America Plate By the Barremian/Aptian (Fig. 2.6e), a plate boundary (NOAM) as spreading ceased in the Gulf of Mexico, and a reorganization in at least the eastern portion of the Proto- Neocomian circum-Gulf carbonate bank was established. Caribbean is required to accommodate the onset of opening Starting in the Berriasian, North America-Gondwana plate in the equatorial Atlantic. The trace of the fracture zone separation occurred solely within the Proto-Caribbean Sea, which originally existed along the Guyana margin must an arm of the Central North Atlantic (Fig. 2.6d). Most have shifted north, possibly as a result of the kinematic margins of the Proto-Caribbean Sea had been formed by adjustments made during the M-21 to M-10 kink in Atlantic rifting, and carbonate and terrigenous shelf deposits accu- fracture zones. Final development of the eastern Bahamas mulated on the subsiding shelves throughout the Creta- basement uplift also probably was associated with this reor- ceous. The Proto-Car ibbean ridge system connected the ganization. central North Atlantic ridge system to plate boundaries in Other important plate boundary reorganizations were 33 the Pacific realm, presumably the backarc spreading centres also about to take place. Engebretson suggested that Far - of central Mexico and offshore Colombia. allon crust may have begun to converge in a more northeast direction relative to North America during Aptian-Albian Early Cretaceous and Aptian-Albian times, which would have relieved the sinistral component Cordilleran orogenesis of strain along the Mexican Cordillera. The opening of the Plate separation continued between North and South equatorial Atlantic was presumably well underway by the America in the Proto-Caribbean, whose passive margin Albian, as evidenced by Albian marine inundation of all sections (Fig. 2.5) continued to develop during thermal equatorial Atlantic rift basins, such that for the first time the subsidence (Fig. 2.6d, e). However, tectonically driven up- South American lithosphere accelerated westward across lift of unknown magnitude in northeastern South America the mantle. The Proto-Caribbean spreading ridge probably may have been caused at the end of the Jurassic and earliest ceased to exist in the late Albian (Figs 2.2, 2.3), but the Cretaceous by a transient shift in the central North Atlantic spreading rate in the central Atlantic increased dramatic ally 48 spreading centre between M-21 and M-1035 . Also at this during the Cretaceous Quiet Period . Thus, South America time, the eastern Bahamas was elevated to sea level so that accelerated from Africa faster in order to 'catch up' with a carbonate bank developed there, either by tectonism re- North America, but in fact both American plates must have lated to the shift in the Atlantic spreading65 or by igneous accelerated westward across the mantle. This would drive the Cordilleran arc systems into a compressional arc con- extrusion which may have been associated with a hot spot 23 trace extending from the Jurassic Florida Volcanic Belt. figuration in the sense of Dewey , in which the overriding Along the Cordillera, the Mexican (Sierra Madre) and plate was actively thrust across the trace of the pre-existing Colombia backarc extensional zones probably continued to trench(es). This process was a chief cause of Sevier and expand as divergence between North and South America Peruvian orogenesis in the continental arc portions of the continued. Deep water sedimentation has been suggested Cordillera, which involved Cordilleran uplift, reduction of from rocks of central Mexico, and ophiolitic rocks occur in the subduction angle, a general eastward shift in the axis of the Juarez Terrane east of Oaxaca18, but it is difficult to volcanism, and backthrusting and associated foredeep basin suggest the width of the basin. In continental portions of development. Colombia and northern Ecuador, there is no record of arc However, the Aleutian-like Antillean-Amaime intra- magmatism during the early and middle Cretaceous, sug- oceanic arc between Chortis/Mexico and southern Ecuador

26 JAMES L. PINDELL

Figure 2.6e. Palaeogeography, Barremian. appears to have flipped its polarity (Fig. 2.6f, g), possibly by the arc was probably over 2,000 km long, the structural the evolution of backthrusting taken to the extreme by the disruption within the arc as it changed convexity could have creation of a new Benioff zone. Aptian/Albian metamor-phic been intense: the arc may have been first shortened (uplift) ages, often from high-pressure minerals, are common around and then extended (subsistence with much rifting) during the Antilles, Tobago, the Villa de Cura complex of the flip. Most of the arc's pre-Albian palaeogeographic Venezuela, the Ruma zone of northern Colombia and the elements were probably rendered beyond recognition, with Amaine terrane of Colombia, and possibly relate to the many 'new' areas of the arc infilling gaps (such as larger orogenesis pertaining to the flip and/or to the onset of west- batholiths) between older rearranged fragments. Likewise, dipping subduction (possible mechanism to elevate seismic sections of the internal Caribbean Sea's lithosphere blueschists toward the surface) on the east side to the arc74. indicate that it underwent significant deformation in mid- Albian metamorphic ages, such as from the Villa de Cura of Cretaceous times, including extrusion and intrusion of ba- Venezuela10, probably pertain to terranes which were pre- salts and dolerites (seismic horizon B") onto and into, served at relatively shallow levels after the flip, such that the pre-existing Pacific-derived sediments and crust of probable metamorphic age was preserved. Other terranes, such as the early Cretaceous and ?Jurassic age. It has been suggested rocks of the northern Cordillera de la Costa Belt of central that the arrival from the Pacific of buoyant B" lithosphere at Venezuela8, probably remained deeper in the forearc or the the west-facing arc helped to cause the flip in polarity51, and trench setting such that they continued to develop metamorphic this may be so, but the age of the B" material (approximately fabrics and ages at younger times and at progressively lesser Albian? to Coniacian) suggests that its extrusion may be a depths. In some locations, the magmatic axis of the arc itself result, rather than a cause, of the flip. shifted during the Albian, for example in the Dominican Although many details of this event remain to be Republic from the Los Ranchos area to the Central worked out, I will assume here that the well-known late Cordillera, possibly as a result of the reversal. Albian to Eocene phase of magmatic activity in many Car- The polarity reversal theoretically should have trans- ibbean arc terranes with Aptian-Albian, possibly amalga- formed the configuration of the intra-oceanic arc from con- mated, metamorphic basements is due to subduction of vex to the west, to generally convex to the east. Given that Proto-Caribbean and Colombian backarc oceanic crust be-

27 The Gulf of Mexico and the Caribbean

Figure 2.6f. Palaeogeography, late Albian. neath the Greater Caribbean arc after a highly orogenic tinued along the margins of the Proto-Caribbean during Aptian-Albian reversal of subduction polarity (see, for ex- thermal subsidence in the absence of plate boundaries (Fig. ample, Fig. 2.6g). The reversal correlates tectonically to the 2.6g). The late Albian drowning of carbonate shelves and Sevier and Peruvian orogenesis in continental areas to rapid landward transgression at most portions of the Proto- the north and south. From late Albian to Santonian times, Caribbean margins may have been enhanced by the death of the arc system may have defined the eastern edge of the Proto-Caribbean spreading ridge. The death of the ridge Farallon crust, but the Santonian onset of subduction at the would reduce in-place stress around the region, due to the Panama-Costa Rica arc isolated the Caribbean lithosphere loss of the ridge push force as it subsided, also allowing the by Campanian time as an independent plate with its own 69 margins to subside. In contrast, arc activity was underway relative motion history for the first time . Once this was along the Greater Antilles, the Amaime-Chaucha Terrane achieved, the east-west component of motion of the (for example, Buga Batholith), and the Mexico/Chortis and Caribbean lithosphere remained approximately in the southern Ecuador/Peruvian margins. mantle reference frame, unable to shift east or west The onset of relative eastward migration of the Cordil- because of its bounding Benioff zones. Continued leran systems led to the progressive closure of the Mexican westward drift of the Americas from Africa produced the and Colombian backarc basins (Fig.2.6g, h). In Mexico, apparent late Cretaceous to Recent eastward migration of closure probably began in the Albian as in the rest of the the Caribbean relative to the Americas, at rates very close Cordillera, but orogenic facies did not commonly appear to the Atlantic spreading rate through time. until the Campanian. This was because the thrust belt would not become emergent until enough shortening had accumu- CARIBBEAN EVOLUTION: PHASE 2; lated to roughly match the amount of extension which had CONSUMPTION OF THE PROTO-CARIBBEAN taken place during the early Cretaceous. To the southeast, SEAWAY BENEATH THE CARIBBEAN PLATE arc-continent collision and northward obduction of the Santa Cruz ophiolite took place alone southern Yucatan Late Cretaceous during the Campanian-Maastrichtian80,94. This was mar ked Non-volcanogenic shelf sedimentation generally con- by the drowning of Cuban shelf carbonates by deeper-water

28 JAMES L. PINDELL

Figure 2.6g. Palaeogeography, Turonian.

Campur carbonates and eventually Sepur Formation fore- Eastward-dipping subduction began in the Panama- deep flysch derived largely from the arc. Similar flysch Costa Rican arc along the Pacific side of B"-affected crust sequences probably flowed eastwards along the Cuban in the Santonian, as indicated by rapid uplift and the onset trench to be accreted to the Cuban thrust belt during its of Campanian volcanogenic sandstone and shallow-water journey toward the Bahamas. In the Antillean segment of carbonate deposition53. The formation of this arc at this the arc, subduction of Proto-Caribbean crust continued particular time was required in order to take up the rapidly throughout the late Cretaceous, but the Campanian was a accelerating Farallon-North America convergence rate of time of uplift, erosion and deformation (for example, in 33 12 about 150 mm/yr . The Caribbean proceded to move at Hispaniola ), probably related to plate interactions accom- about 30 mm/yr70, so that the new arc, which defined the modating the Proto-Caribbean bottleneck between Yucatan Caribbean Plate for the first time, took up the difference of and Colombia. over 100 mm/yr. I suggest that the arc stretched from west In Colombia, the Amaime-Chaucha Terrane had been of Chortis, southeastwards to southern Ecuador. The Ecua- diachronously accreted by eastward-vergent thrusting onto dorian site for the trench-trench-transform fault (TTT) triple the Central Cordilleran passive margin by the end of the junction is indicated by the boundary between voluminous Campanian, which (1) thermally reset many older radio- and nearly absent Andean volcanism at this time to the south genic systems in the Central Cordillera, producing the ap- and north, respectively: to the north, Caribbean plate con- pearance of Cretaceous magmatism there, and (2) drove vergence with the Andes was only approximately 20 mm/yr, foredeep basinal subsidence (Colon, Umir Formations) east whereas to the south the Farallon Plate's convergence rate of the Central Cordillera by tectonic loading from the 74 with the Andes was over 100 mm/yr. During the remainder west . Toward the end(?) of the Campanian, continued of the Cretaceous and into the Tertiary, the triple junction convergence began to occur at an east-dipping trench out- appears to have migrated northwards and eventually into board of the Amaime Terrane, leading to the progressive southern Colombia (Fig. 2.6i, j, k), at the rate of the north- accretion of Caribbean B" and other sedimentary materials ward component of motion in this area between the Carib- into the Western Cordillera Belt during Maastrichitian and bean and South American plates, with a corresponding early Paleogene times (Fig. 2.6h). increase in magmatism in the wake. However, this level of

29 The Gulf of Mexico and the Caribbean

Figure 2.6h. Palaeogeography, Campanian.

onshore Cuba may represent mainly the forearc complex if detail cannot be stated with confidence without further rifting occurred nearly along the arc axis. A possible reason work. Finally, in regard to the Cretaceous, the problem of that Cuba lacks latest Cretaceous -Paleogene magmatic having both the Chortis Block as well as the Caribbean Plate rocks is that the presently subaerial portion of Cuba was too move synchronously eastwards relative to the Americas still close to the former trench: the arc-trench gap should have appears to be best solved by northward underthrusting or been greater, and plutons and volcanics of that age may exist subduction along the Lower Nicaraguan Rise69. The char - offshore to the south. The opening of the Yucatan basin is acter of the Rise is not unlike a submarine, oceanic crust- defined by the three-plate system (NOAM-Caribbean- 22 Cuba); trends and magnitudes of the relative motions may bearing accretionary prism and volcanism in the Upper 69 Nicaraguan Rise to the north continued at least into the be calculated by vector completion . Collision between the Paleocene. The Chortis Block must have moved relatively Greater Antilles and the Bahamas began in the Paleocene, eastwards, while the Caribbean must have moved eastnorth- although subduction accretion packages occur onshore east, thereby producing convergence between the two at the Cuba which formed during the late Cretaceous and included late Cretaceous orogenic sediments probably derived from Nicaraguan Rise. Yucatan, confusing the definition of the exact age of the Cenozoic onset of the Cuba-Bahamas collision. A Paleogene age is also generally accepted for the The Caribbean Plate continued migration relatively 89 eastnortheastwards, subducting oceanic crust of the Proto- opening of the Grenada Basin . North-south extension best Caribbean. The Yucatan Basin opened by intra-arc spread- explains the east-west magnetic pattern and orientation of ing and extreme attenuation of Greater Antilles arc crust78 normal faults in and around the basin, as well as the sharp in a three-plate system (Fig. 2.6j, vector inset) between southeast boundary of the Aves Ridge (transform faulted?). Cuba, the Cayman Ridge and North America69. The bulk of Dextral oblique subduction of the Proto-Caribbean crust, the Cuban portion of the Greater Antilles magmaiic arc is dextral transform drag along South America, and subduc - split between the basement of the Cayman Ridge and Cuba's tion zone rollback of the Jurassic oceanic crust along the northern South America passive margin may have combined southern arc belt. The Cretaceous volcanic assemblages of

30 JAMES L. PINDELL

Figure 2.6i. Palaeogeography, Maastrichtian. to drive the Andaman Sea-type intra-arc extension. The to the Lesser Antilles arc, beginning in the Eocene in the opening of both the Yucatan and the Grenada intra-arc northern Lesser Antilles, but possibly not until the Oligo- basins was the mechanism by which the Caribbean Plate cene or early Miocene in the south. This diachroneity prob- accommodated the shape of the Proto-Caribbean basin. It ably relates to the north-south opening of the Grenada Basin was apparently easier to rift the arc complexes (driven by and a consequently lesser subduction rate (component of rollback of existing Benioff zones) than to tear railroad eastward migration) in the south during the Eocene. transforms in the Jurassic Proto-Caribbean oceanic crust. In the middle to late Eocene, just after the Antilles-Ba- To the west, the Yucatan block prevented simple east- hamas collision, east-northeast migration of the Caribbean northeastward motion of Chortis and the Nicaraguan Plate relative to North America continued along a new plate Rise/Jamaica with the rest of the Caribbean Plate, and boundary system which became the northern Caribbean compression was consequently set up between Chortis and plate boundary zone (NCPBZ). The Cayman Trough nucle- the Caribbean Plate. Chortis, the Nicaraguan Rise, and ated as a pull-apart basin between Yucatan and Jamaica. Jamaica were internally deformed during the Paleogene Cuba, the Cayman Ridge and the Yucatan Basin were left (Wagwater and Montpelier Troughs in Jamaica; rifts of the 54 as apart of the North American Plate by the development of Nicaraguan Rise ). the NCPBZ. The middle Eocene was marked by the termination of To the south, subduction of Caribbean crust beneath Bahamian-Antillean collision and the onset of platform northwest Colombia produced an accretionary prism of deposition in Cuba (overlap assemblage). Extension in the Andean, Incaic phase orogenic sediments (San Jacinto Belt) Yucatan Basin ceased as Cuba came to rest against the which grew until the Miocene31,32. Progressive development Bahamas. The reconstruction of Figure 2.6k is constrained of the early Barbados accretionary prism88 was due to by: restoring 1,050 km of offset in the Cayman Trough; the southeastward migration relative to South America of the alignment of late Cretaceous-Eocene, subduction-related terrane to the east of the Grenada Basin (Grenada Terrane) plutons throughout the Greater Antilles and Aves Ridge; during the opening of the Grenada Basin (three-plate sys- and sedimentary facies changes across the Greater An- 69 tem; Fig. 2.6j). Quartzose sands of Barbados and the tilles . Volcanism shifted eastwards from the Aves Ridge Piemontine nappes of central Venezuela were accreted to

31 The Gulf of Mexico and the Caribbean

Figure 2.6j. Palaeogeography, Paleocene. the complex in the Eocene prior to emplacement onto the Rico into contact with central Puerto Rico along the 'Eocene shelf. These probably originated from western and central Tectonic Belt'. Venezuela, where Precambrian acidic massifs were exposed In the developing southern Caribbean plate boundary 11,87 at that time, and from the peripheral bulge ahead of the zone (SCPBZ) , eastward migration of the Caribbean Venezuela foredeep basin. Plate progressively lengthened the zone of Caribbean-South Post-Eocene subduction of Proto-Caribbean crust is American interaction (Fig. 2.61). Nappes have been em- indicated by subduction-related magmatism in the Lesser placed southeastwards onto the Venezuela margin since the Antilles. In the NCPBZ, the Cayman Trough progressively Paleocene; initiation of thrusting upon the autochthon be- opened by seafloor spreading at the Mid -Cayman Spreading comes progressively younger to the east. Subsidence of the Centre, which linked transform faults connecting to the Venezuela shelf near Maracaibo was Paleocene-early Eo- 95 Middle America and Lesser Antilles subduction zones (Fig. cene and subsidence in the Eastern Venezuela Basin was 91 2.6j, k). These transforms have changed location through Miocene-Pliocene . A transform fault between the Carib- time, forming anastomosing fault systems across central bean Plate and the obducted terranes on South America Guatemala in the west and across Hispaniola/Puerto Rico in basement developed progressively after thrust emplace- the east. Large-offset transcurrent motions between blocks ment, and assisted with continuing strike-slip offset of the of Hispaniola are indicated by incompatible Tertiary sedi- Caribbean, but not the obducted terranes, after emplace- 64 84 mentary facies which are presently juxtaposed across fault ment. The Falcon and Cariaco Basins are two examples zones. The primary offset during the late Eocene and Oligo- of Oligocene-early Miocene and Miocene-Recent, respec- cene occurred along faults of the northern San Juan Basin, tively, transtensional basins that developed in previously juxtaposing the San Juan block with the Cordillera Central- overthrust areas. I note that both opened only after the Massif du Nord arc by the early Miocene, as indicated by development of the Grenada Basin, when the relative mo- the flooding of arc-derived elastics into the San Juan Basin tion vector was extremely oblique. at that time20,60. Motion on eastward extensions of this fault Motion through Hispaniola during the early to middle system separated Puerto Rico from central Hispaniola, and Miocene (Fig. 2.6m) continued along the northern San Juan may have brought the Bermeja area of southwest Puerto Basin, and possibly along the south flank of Sierra Neiba,

32 JAMES L. PINDELL

Figure 2.6k. Palaeogeography, middle Eocene. but the main locus of motion became the Oriente Fault Cordillera, marking the suture zone within the Western between Cuba and Hispaniola at about 20-25 Ma. This was Cordillera rather than at the Atrato Basin; and (4) shows that responsible for the present separation of the two islands. By arc magmatism, which did not begin in Colombia until the the late Miocene, convergence and uplift in Sierra Neiba had middle? Miocene, pertains to subduction of the Cocos or structually separated the San Juan and Enriquillo Basins. Nazca Plates, rather than of the Caribbean Plate. In any case, The logical continuation of this system is the Muertos 50 the progressive collision hindered and eventually blocked Foldbelt , which has become a south-vergent overthrust circulation between the Caribbean and Pacific45, and was a zone probably since Miocene time. major cause of northern Andean compression and uplift, In the south, the North Venezuelan and Piemontine helping to drive the Maracaibo Block northwards from the nappes reached final emplacement onto the Venezuelan Eastern Colombian Cordillera. This escape produced the shelf, overthrusting the Oligocene-early Miocene foredeep South Caribbean Foldbelt north of the Maracaibo block and basin (Roblecito Formation). Out in the Pacific, spreading offshore terranes 25, and development of the Panama Oro- was initiated at the Galapagos spreading centre; its relation- cline produced the North Panama Foldbelt52. The arc terra- ship to the Caribbean plate boundary circuit is unclear nes of Cuba and Hispaniola continued to separate by because its eastern portions have already been subducted. transform motion along the present-day Oriente Fault. Com- The Panama arc, a part of the Caribbean Plate, has often pression has continued in the Sierra Neiba/Enriquillo sys- been shown as having collided with Colombia in the Mio- tems, contributing to the present-day complexity of the cene. In Figure 2.6k-n, I show a speculative, more prolonged NCPBZ. history of interaction, which: (1) considers slower relative Miocene to Recent deformation around the Caribbean motion rates farther south such that Panama was always was and is common and very strong. This Neo-Caribbean closer to Colombia than previously thought (16 mm/yr for phase of deformation17,69 results from the continued drift of Neogene, 24 mm/yr for Paleogene); (2) portrays the Panama the Americas westward past the Caribbean, and from the Orocline as a slowly developing feature as plate conver- general state of compression which can be related to at least gence continued; (3) allows for the east-vergent obduction three causes. First, North-South American relative motion of the Panamanian arc, or Choco Terrane, onto the Western vectors (Fig. 2.2) show convergence during the Neogene,

33 The Gulf of Mexico and the Caribbean

Figure 2.6l. Palaeogeography, early Oligocene. which constricts the Caribbean Plate. Second, the restrain- REFERENCES ing bend in the Oriente-Puerto Rico Trench transform fault 1 northeast of the Dominica Republic 13 constricts the east- Algar, S.T. 1993. Structural, stratigraphic, and thermo- ward migration of the north-central Caribbean and is respon- chronologic evolution of Trinidad. Unpublished Ph.D. thesis, Dartmouth College, Hanover. sible for much transpression in Hispaniola. The eastern part 2 of Hispaniola, which has already passed this bend, has Algar, S.T. & Pindell, J.L. 199la. Structural development subdued topography relative to the western part. Third, the of the Northern Range of Trinidad, and implications for northeastward migration, relative to the Guyana Shield, of the tectonic evolution of the southestera Caribbean: in the Andean Cordilleran Terranes has induced a compression Gillezeau, K. A. (ed.), Transactions of the Second Geo- upon the Caribbean Plate to the northnorthwest, because the logical Conference of the Geological Society ofTrin- Caribbean Plate possesses an eastward component of mo- dad and Tobago, Port-of-Spain, Trinidad, April 3-8, 1990, 6-22. tion relative to the Guyana Shield that is slightly greater than 3 that of the Cordilleran Terrane26. In the southeast Carib- Algar, S.T. & Pindell, J.L. 1991b. Stratigraphy and sedi- bean, where the Caribbean and South American plates mentology of the Toco region of the Northern Range of nearly come into contact, the relative motion of the two has Trinidad: in Gillezeau, K.A. (ed.), Transactions of the been slightly north of east since the late Miocene4, but was Second Geological Conference of the Geological Society more convergent (transpressional to the eastsoutheast) in the ofTrindad and Tobago, Port-of -Spain, Trinidad, April 3- early and middle Miocene. A fourth cause, which pertains 8, 1990,56-69. 4 mainly to Colombia, is that the crust of the Caribbean Plate Algar, S.T. & Pindell, J. 1993. Structure and deformation is buoyant and resists subduction, as attested to by the history of the Northern Range of Trinidad and adjacent Andean orogenesis which has occurred in the absence of areas. Tectonics, 12, 814-829. 5 volcanism throughout the Cenozoic. Anderson, T.H., Burkart, B., Clemons, R.E., Bohnenber- ger, O.K. & Blount, D.N. 1973. Geology of the Cuchu- ACKNOWLEDGEMENTS—I thank John Dewey, Walter Pitman, Edward mantanes, northwestern Guatemala. Geological Robinson, Sam Algar and Johan Erikson for their input on various elements Society of America Bulletin, 84, 805-826. of this review.

JAMES L. PINDELL

Figure 2.6m. Palaeogeography, early Miocene.

6 Anderson, T.H. & Schmidt, V.A. 1983. The evolution of Venezuela margin. American Association of Petroleum Middle America and the Gulf of Mexico-Caribbean Sea Geologists Memoir, 34, 347-358. region during Mesozoic time. Geological Society of 12Bowin, C. 1975. The geology of Hispaniola: in Nairn, America Bulletin, 94, 941-966. A.E.M. & Stehli, F.G. (eds), The Ocean Basins and 7 Ave Lallemant, H.G. & Oldow, J.S. 1988. Early Mesozoic Margins. Volume 3. The Gulf of Mexico and the Carib- southward migration of Cordilleran transpressional ter- bean, 501-550. Plenum, New York. ranes. Tectonics, 7,1057-1075. 13 8 Bracey, D.R. & Vogt, P.R. 1970. Plate tectonics in the Ave Lallemant, H.G. & Sisson, V.B. 1993. Caribbean- Hispaniola area. Geological Society of America Bulle- South Americ an interactions: constraints from the Car- tin, 81,2855-2860. ibbean de la Costa Belt, Venezuela: in Pindell, J.L. 14Buffler, R.T. & Sawyer, D.S. 1985. Distribution of crust (ed.), Mesozoic and Early Cenozoic Development of the and early history, Gulf of Mexico Basin. Transactions Gulf of Mexico Caribbean Region. Transactions of the of the Gulf Coast Association of Geological Societies, 13th Annual Research Conference, Gulf Coast Section, 35,333-344. SEPM Foundation, 211-220. 15 9 Buffler, R.T. & Thomas, W. (in press). Crustal structural Barr, K.W. 1963. The Geology of the Toco District, Trini- and tectonic evolution of the southeastern margin of dad, W.L Overseas Geological Surveys, HMSO, Lon- North America and the Gulf of Mexico basin: in Speed, don. 10 R. (ed.), The geology of North America, Volume CTV-1, Beets, DJ., Maresch, W.V., Klaver, G.Th., Mottana, A., Phanerozoic Evolution of North American Continent- Bocchk), R., Beunk, F.F. & Monen, H.P. 1984. Mag- Ocean Transition. Geological Society of America, matic rock series and high-pressure metamorphism as Boulder. constraints on the tectonic history of the southern Car- 16Burke, K. 1988. Tectonic evolution of the Caribbean. ibbean. Geological Society of America Memoir, 162, Annual Review of Earth and Planetary Sciences, 16, 95-130. 11 201-230. Biju-Duval, B, Mascle, A., Rosales, H. & Young, G. 17Burke, K., Grippi, J. & Sengor, A.M.C. 1980. Neogene 1983. Episutural Oligo-Miocene basins along the north structures in Jamaica and the tectonic style of the north-

35 The Gulf of Mexico and the Caribbean

Figure 2.6n. Palaeogeography, late Miocene.

ern Caribbean plate boudary zone. Journal of Geology, The continental crust and its mineral deposits. Special 88,375-386. Paper of the Geological Association of Canada, 20, 18 Campa, M. & Coney, PJ. 1983. Un modelo tectonic de 553-573. Mexico y sus relaciones con America del Norte, Amer- 24Dewey, J.F. 1982. Plate tectonics and the evolution of the ica del Sur y el Caribe. Revista del Instituto Mexicano British Isles. Journal of the Geological Society of Lon- delPetroleo,15,6-l5. don, 129,371-412. 19 Chevalier, Y., Stephan, J.-F., Darboux, J-R., Gravelle, M., 25Dewey, J.F. & Pindell, J.L. 1985. Neogene block tectonics Bellon, H., Bellizzia, A. & Blanchet, R. 1988. Obduc- of Turkey and northern South America: continental tion et collision pre-Tertiaire dons les internes de la applications of the finite difference method. Tectonics, Chain Caraibe venezuelienne, sur le transect fle de 4,71-83. Margarita-Pennisule d'Araya. Compte Rendu de la 26Dewey, J.F. & Pindell, J.L. 1986. Neogene block tectonics Academie des Sciences de Paris, serie II, 307, 1925- of Turkey and northern South America: continental 1932. applications of the finite difference method—reply. 20 Cooper, J.C. 1983. Geology of the Fondo Negro Basin and Tectonics, 5, 703-705. adjacent areas, Dominican Republic. Unpublished 27Dickinson, W.R & Coney, PJ. 1980. Plate tectonics M.S. thesis, State University of New York at Albany. constraints on the origin of the Gulf of Mexico: in 21 Damon, P. & Coney, P. 1983. Rate of movement Pilger, R.H. (ed.), The Origin of the Gulf of Mexico and of nuclear Central America along the coast of the Early Opening of the Central Atlantic, 27-36. Lou- Mexico during the last 90 Ma. Geological Society of isiana State University, Baton Rouge, America, Abtracts with Programs, 15, 553. 28Dolan, J.F., Mann, P., Monechi, S., de Zoeten, R., 22 Dengo, G. & Case, J.E. (eds). 1990. The Geology of North Heubeck, C., & Shiroma, J. 1991. Sedimentologic, America, Volume H, The Caribbean region. Geological stratigraphic, and tectonic synthesis of Eocene-Mio- Society of America, Boulder. cene sedimentary basins, Hispaniola and Puerto Rico. 23 Dewey, J.F. 1980. Episodicity, sequence and style at Geological Society of America Special Paper, 262, convergent plate boundaries: in Strangway, D.W. (ed.), 217-240.

36 JAMES L. PINDELL

29Donnelly, T.W. 1989. Geologic history of the Caribbean 42James, K. 1990. The Venezuelan Hydro habitat: in and Central America: in Bally, A.W. & Palmer, A.R. Brooks, J. (ed.), Classic Petroleum Provinces. Geologi (eds), The Geology of North America—An Overview, cal Society of London Special Publication, 50, 9-36. 43 299-321. Geological Society of Americ a, Boulder. Jansa, L.F. & Weidmann, J. 1982. Mesozoic -Cenozoic 30Duncan, R A. & Hargraves, R.B. 1984. Plate tectonic development of the eastern North American and north evolution of the Caribbean region in the mantle refer- west African continental margins: a comparison: in von ence frame. Geological Society of America Memoir, Rad, U., Hinz, K., Sarnthein, M. & Siebold, E. (eds), 162, 81-84. Geology of the Northwest African Margin, 215-269. 31 Duque-Caro, H. 1979. Major structural elements and evo- Springer-Verlag, New York. lution of northwest Colombia. American Association 44Johnson, C.C. & Kauffman, E.G. 1989. Cretaceous radis- of Petroleum Geologists Memoir, 29, 329-351. tid paleobiogeography of the Caribbean Province. Ab- 32 Duque-Caro, H. 1984. Structural style, diapirism, and stracts, 12th Caribbean Geological Conference. St. accretionary episodes of the Sinu-San Jacinto Terrane, Croix, U.S. Virgin Islands, 7th-llth August, 85. southwestern Caribbean borderland. Geological Soci- 45Keigwin, L.D., Jr. 1978. Pliocene closing of the Isthmus ety of America Memoir, 162,303-316. of Panama, based on biostratigraphic evidence from 33 Engebretson, D.C. 1982. Relative motions between oce- nearby Pacific Ocean and Caribbean sea cores. Geol- anic and continental plates in the Pacific basin. Un- ogy, 6, 630-634. published Ph.D. thesis, Stanford University, California. 46 34 King, P.B. 1969. Tectonic Map of North America, scale Erikson, J.P. 1992. Northeastern Venezuela's Jurassic 1:5,000,000. United States Geological Survey, Reston, through Eocene passive margin, Hispaniola's Neogene 47 Virginia. Cibo Basin, and their histories and causes of evolution. Klitgord, K.D., Popenhoe, P. & Schouten, H. 1984. Florida: Unpublished Ph.D. thesis, Dartmouth College, Hano- a Jurassic transform plate boundary. Journal of ver. Geophysical Research, 89, 7753-7772. 35 48 Erikson, J.P. & Pindell, J.L. (in press). Cretaceous-Eocene Klitgord, K. & Schouten, H. 1986. Plate kinematics of the passive margin relative sea level history, sequence stra- central Atlantic: in Tucholke, B.E. & Vogt, P.R. (eds), The tigraphy, and the tectonic eustatic causes of strati- Geology of North America, Volume M, The Western graphic development in northeastern Venezuela SEPM Atlantic Region, 351-378. Geological Society of Special Publication. America, Boulder. 36 49 Erikson, J., Pindell, J.L. & Larue, D.K. 1990. Mid-Eo- Ladd, J.W. 1976. Relative motion of South America with cene-early Oligocene sinistral transcurrent faulting in respect to North America and Caribbean tectonics. Puerto Rico associated with formation of the northern Geological Society of America Bulletin, 87, 969-976. Caribbean plate boundary zone. Journal of Geology, 50Ladd, J.W. & Watkins, J.S. 1978. Tectonic development 98,365-384. 37 of trench-arc complexes on the northern and southern Gonzalez de Juana, C., Arozena, J.A. & Picard Cardillat, margins of the Venezuelan Basin: in Watkins, J.S., X. 1980. Geologiade Venezuela y sus Cuencas Petro- Montadert, L. & Dickerson, P.W. (eds), Geological and liferas. Ediciones Foninzes, Carcas. 38 Geophysical Investigations of Continental Margins. Hall, S.A., Najmuddin, I. & Buffler, R.T. (in press). American Association of Petroleum Geologists Mem- Contraints on the tectonic development of the Gulf of oir, 29, 363-371. Mexico provided by magnetic anomaly data over the 51Livacarri, R.F., Burke, K. & Sengor, A.M.C. 1981. Was deep Gulf. Journal of Geophysical Research. the Laramide Orogeny related to subduction of an oce- 39Hilst, R.D. van der. 1990. Tomography with P, PP and pP anic plateau? Nature, 189,276-278. delay-time data and the three-dimension mantle struc- 52Lu, R.S. & McMillen, K.J. 1983. Multichannel seismic ture below the Caribbean region. Geological Ultraiecu- survey of the Colombia Basin and adjacent margin: in tina, Mededelingenvan de Faculteit Aardwetenschap- Watkins, J.S. & Drake, C.L. (eds), Studies in Continental pen der Rijksuniversiteit te Utrecht, 67, 250pp. 40 Margin Geology. American Association of Petro- Imlay, R.W. 1980. Jurassic paleobiogeography of the leum Geologists Memoir, 34, 395-410. conterminous United States in its continental setting. 53 Lunberg, N. 1983. Development of forearcs of intrao- U.S. Geological Survey, Professional Paper, 1062, 1- ceanic zones. Tectonics, 2, 51-61. 134. 54 41 Mann, P. & Burke, K. 1984a. Cenozoic rift formation in the Irving, E.M. 1975. Structural evolution of the northernmost Andes, Colombia. U.S. Geological Survey, northern Caribbean. Geology, 12,732-736. 55Mann, P. & Burke, K. 1984b. Neotectonics of the Caribbean. Professional Paper, 846, 1-47. Reviews of Geophysics and Space Physics, 22,

37 The Gulf of Mexico and the Caribbean

309-362. 68Pindell, J.L. 1991. Geological rationale for hydrocarbon 56Mann, P., Burke, K. & Matumoto, T. 1984. Neotectonics exploration in the Caribbean and adjacent regions. of Hispaniola: plate motion, sedimentation, and seis- Journal of Petroleum Geology, 14, 237-257. micity at a restraining bend. Earth and Planetary Sci- 69Pindell, J.L. & Barrett, S.F. 1990. Geologic evolution of ence Letters, 70,311-324. the Caribbean region: a plate-tectonic perspective: in 57Maresch, W.V. 1974. Plate tectonics origin of the Carib- Dengo, G. & Case, J.E. (eds), The Geology of North bean mountain system of northern South America: dis- America, Volume H, The Caribbean Region, 405-432. cussion and proposal. Geological Society of America Geological Society of America, Boulder. 70 Bulletin, 85, 669-682. Pindell, J.L., Cande, S.C., Pitman, W.C., Rowley, D.B., 58Marton, G. & Buffler, R.T. 1993. The southern Gulf of Dewey, J.F., LaBrecque, J. & Haxby, W. 1988. A Mexico in the framework of the opening of the Gulf of plate-kinematic framework for models of Caribbean Mexico Basin: in Pindell, J.L. (ed.), Mesozoic and evolution. Tectonophysics, 155, 121-138. Early Cenozoic Development of the Gulf of Mexico 71Pindell, J.L., Drake, C.L. & Pitman, W.C. 1991. Prelimi- Caribbean Region. Transactions of the 13th Annual nary assessment of a Cretaceous to Paleogene Research Conference, Gulf Coast Section, SEPM Atlantic passive margin, Serrania del Interior and Foundation, 51-68. Central Ranges, Venezuela/Trinidad. Abstracts, 59Mattson, P.H. & Pessagno, E.A., Jr. 1979. Jurassic and American Association of Petroleum Geologists early Cretaceous radiolarians in Puerto Rican ophiolite; Annual Meeting, Dallas, 190. 72 tectonic implications. Geology, 7,440-444. Pindell, J.L. & Dewey, J.F. 1982. Permo-Triassic recon- 60Michael,R.C. 1979. Geology of the south-central flank of struction of western Pangea and the evolution of the the Cordillera Central and the adjacent portions of the Gulf of Mexico/Caribbean region. Tectonics, 1, 179- San Juan Valley between Rio San Juan and Rio Yaca- 212. hueque, Dominican Republic. Unpublished M.S. thesis, 73Pindell, J.L. & Draper, G. 1991. Stratigraphy and tectonic George Washington University, Washington D.C. development of the Puerto Plata area, northern Domini- 61Molnar, P. & Sykes, L.R. 1969. Tectonics of the Carib- can Republic. Geological Society of America Special bean and Middle America regions from focal mecha- Paper, 262, 97-114. nisms and seismicity. Geological Society of America 74Pindell, J.L. & Erikson, J.P. (in press). The Mesozoic Bulletin, 80, 1639-1684. passive margin of northern South America: in Vogel, 62 Montgomery, H., Pessagno, E.A., Jr. & Munoz, I.M. A. (ed.), Cretaceous Tectonics in the Andes. Vieweg 1992. Jurassic (Tithonian) radiolaria from La Desirade Publishing, Weisbaden. 75 (Lesser Antilles): preliminary paleontology and tec Pindell, J.L., Erikson, J. & Algar, S. 1991. The relation- tonic implications. Tectonics, 11,426-432. ship between plate motions and sedimentary basin de- 63 Morris, A.E., Taner, I., Meyerhoff, H.A. & Meyerhoff, velopment in northern South America: from a A. A. 1990. Tectonic evolution of the Caribbean region; Mesozoic passive margin to a Cenozoic eastwardly- alternative hypothesis: in Dengo, G. & Case, J.E. (eds), progressive transpressional orogen: in Gillezeau, K.A. The Geology of North America. Volume H. The Carib (ed.), Transactions of the Second Geological Confer- bean Region, 433-457. Geological Society of America, ence of the Geological Society of Trinidad and Tobago, Boulder. Port-of-Spain, Trinidad, April 3-8, 1990, 191-202. 64Muessig, K.W. 1984. Structure and Cenozoic tectonics of 76Pindell, J.L., Robinson, E. & Dewey, J.F. (in press). the Falcon Basin, Venezuela and adjacent areas. Geo Paleogeographic evolution of the Gulf of Mexico/ logical Society of America Memoir, 162, 217-230. Caribbean Region. SEPM Special Publication. 77 65Pindell, J.L. 1985a. Alleghenian reconstruction and the Rodgers, D.A. 1984. Mexia and Talco fault zones, east subsequent evolution of the Gulf of Mexico, Bahamas Texas: comparison of origins predicted by two tectonic and Proto-Caribbean Sea. Tectonics, 4,1-39. models: in Presley, M.W. (ed.), The Jurassic of East 66Pindell, J.L. 1985b. Plate-tectonic evolution of the Gulf of Texas. Transactions of the East Texas Jurassic Explo- Mexico and Caribbean Region. Unpublished Ph.D. the ration Conference, 23-31. East Texas Geological Soci- sis, University of Durham, Durham. ety, Tyler. 67 78 Pindell, J.L. 1990. Arguments for a Pacific origin of the Rosencrantz, E. 1990. Structure and tectonics of the Yu- Caribbean Plate: in Larue, D.K. & Draper, G. (eds), catan Basin, Caribbean Sea, as determined from seis - Transactions of the 12th Caribbean Geological Con mic reflection studies. Tectonics, 9, 1037-1059. 79 ference, St. Croix, U.S. Virgin Islands, 7th-llth August, Rosencrantz, E., Ross, M. & Sclater, J.G. 1988. Age and 1989, 1-4. Miami Geological Society, Florida. spreading history of the Cayman Trough as determined

38 JAMES L. PINDELL

from depth, heat flow, and anomalies. Journal of Geo- 87Silver, E.A., Case, J.E. & Macgillavry, H.J. 1975. Geo- physical Research, 93, 2141-2157. physical study of the Venezuelan boreland. 80Rosenfeld, J.H. 1993. Sedimentary rocks of the Santa Geological Society of America Bulletin, 86, 213-226. Cruz ophiolite, Guatemala—a Proto-Caribbean his- 88Speed, R.C. 1985. Cenozoic collision of the Lesser An- tory: in Pindell, J.L. (ed.), Mesozoic and Early Ceno- tilles Arc and continental South America and origin zoic Development of the Gulf of Mexico Caribbean of the El Pilar Fault. Tectonics, 4, 40-70. Region. Transactions of the 13th Annual Research 89Speed, R.C. etaL 1984. Lesser Antilles Arc and adjacent Conference, Gulf Coast Section, SEPM Foundation, terranes. Ocean Margin Drilling Program, Regional 173-180. 81 Atlas Series, Atlas 10. Marine Science International, Rowley, D.B. & Pindell, J.L. 1989. End Paleozoic -early Woods Hole. Mesozoic western Pangean reconstruction and its im- 90Stoffa, P.L., Mauffret, A., Truchan, M. & Buhl, P. 1981. plications for the distribution of Precambrian and Pa- "Sub-B" layering in the southern Caribbean: the Aruba leozoic rocks around Meso-America. Precambrian gap and Venezuela Basin. Earth and Planetary Science Research, 42, 411-444. 82 Letters, 53, 131-146. Salvador, A. & Green, A.G. 1980. Opening of the Carib- 91Vierbuchen, R.C. 1984. The geology of the El Pilar fault bean Tethys, geology of the Alpine chain born of the zone and adjacent areas in northeastern Venezuela. Tethys. Memoir of the 26th International Geological Geological Society of America Memoir, 162,189-212. Congress. Bureau de Recherches Geological et Min- 92Vogt, P.R., Anderson, C.N. & Bracey, D.R. 1971. erologie, 115, 224-229. 83 Mesozoic magnetic anomalies, seafloor spreading, Schlager, W. et al 1984. Deep Sea Drilling Project, Leg and geomagnetic reversals in the southwestern north 77, southeastern Gulf of Mexico. Geological Society of Atlantic. Journal of Geophysical Research, 76,4796- America Bulletin, 95, 226-236. 84 4823. Schubert, C. 1982. Origin of Car iaco Basin, southern 93Wadge, G. & Burke, K. 1983. Neogene Caribbean plate Caribbean Sea. Marine Geology, 47, 345-360. 85 rotation and associated Central American tectonic Shaw, P.R. & Cande, S.C. 1990. High-resolution inver- evolution. Tectonics, 2, 633-643. sion for South Atlantic plate kinematics using joint 94Wilson, H.H. 1974. Cretaceous sedimentation and oro- altimeter and magnetic anomaly data. Journal of Geo- geny in nuclear Central America. American Associa- physical Research, 95, B2625-B2644. 86 tion of Petroleum Geologists Bulletin, 58, 1348-1396. Sheridan, R.E., Crosby, J.T., Kent, K.M., Dillon, W.P. & Paull, 95Zambrano, E., Vasquez, E. , Duval, B., Latreille, M. & C.K. 1981. The geology of the Blake Plateau and Bahamas. Coffinieres, B. 1972. Paleogeographic and petroleum Canadian Society of Petroleum Geologists synthesis of western Venezuela. Editions Technip, Memoir, 1, 487-502. Paris.

CHAPTER 3

The Caribbean Sea Floor

THOMAS W. DONNELLY

Department of Geological Sciences and Environmental Studies, State University of New York at Binghamton, P.O. Box 6000, Binghamton, New York 13902-6000, U.SA.

INTRODUCTION arc of the Lesser Antilles. The depth of water on the western margin ranges from zero (Aves Island) to a more repre- THE FLOOR of the Caribbean Sea (Figs 3.1,3.2) is essen- sentative 1-2 km. In the southern half of the ridge the tially oceanic. It consists of several basinal areas (from west Grenada Basin is a basinal area with a water depth of nearly to east, the Yucatan, Colombian and Venezuelan Basins), 3km. separated by shallower areas (the Nicaraguan Rise and The origin of the Aves Ridge is obscure, but rocks of Beata Ridge). The Cayman Trough is a narrow, very deep island-arc affinity have been dredged and drilled at several linear basin immediately north of the Nicaraguan Rise and places. Donnelly13 speculated that the ridge represents two is notable for being the site of contemporary sea-floor extinct island arcs of late Cretaceous to Paleogene age, one spreading. The Caribbean sea floor has been extensively on the western margin facing west and one on the eastern described in the recent "Caribbean Region" volume of the margin facing east These arcs have been sundered by late Geological Society of America's Decade of North American Cretaceous to Paleogene eastward movement of the Carib- 9 Geology series . Papers in that volume which are especially bean plate. The Neogene Lesser Antilles island arc is built 7 32 pertinent to the present discussion include chapters 2 , 9 , upon diverse sundered fragments of these older arcs. 1041 and 1316. Extensive reference is also made here to the results of the Deep Sea Drilling Project (DSDP), notably Venezuelan Basin 20 Leg 15 (1970-1971 ). The Venezuelan Basin is the easternmost, the most The floor of the main part of the Caribbean—that is, the thoroughly studied, and the simplest of the basinal areas of Venezuelan and Colombian Basins—is underlain by a vast the Caribbean. Because it has been studied more completely plateau basalt which is not typical of oceanic crust, but than the other basins, and also because it was drilled in which has close analogues elsewhere in the world ocean. several places during DSDP Legs 4 and 15, it will be useful The origin and evolution of this basalt is central to the to consider it in some detail before proceeding to the remain- history of the Caribbean region and will be the main focus ing areas of the Caribbean. of this chapter. The Venezuelan Basin is mainly between 4 and 5 km deep. It is slightly shallower in the centre than on the MORPHOLOGY OF THE CARIBBEAN BASINS northern and southern margins, where it is bounded by narrow troughs with turbidite plains. The trough on the north The following passage briefly describes the morphological is called the Muertos Trough; the one on the south is less units of the Caribbean from east to west. Figures 3.1 and 3.2 well defined and has no commonly accepted name. The locate these features. northern and southern boundaries show clear evidence of young subduction41 of the crust of the basin beneath the Aves Ridge adjacent land areas. The Aves Ridge is a complex elevated submarine area On the east the Venezuelan Basin is bounded by the with a virtually straight north-south western margin and an Aves Ridge. The tectonic nature of this boundary has not eastern margin that merges with the curved volcanic inner been clarified, because the nature and history of the Aves

41 The Caribbean Sea Floor

Figure 3.1. (A) Map of the Caribbean showing major named features within the basinal portion. Filled circles with numbers are drilling sites (DSDP Leg 15) that reached the basaltic crust; open circles are DSDP sites not reaching basement. Line shows crustal cross section of (B). (B) East-west crustal cross-section at 15° latitude, bent to cross Central America at about S45°W. The 'Caribbean basalt/sediment layer' is the upper part of the thickened Caribbean crust; the sediment portion of this layer is conjectured. This figure has been adapted, with minor interpretive revisions, from Case et al7.

Ridge is not fully clear. Near the boundary there is a pattern island-arc affinity from the western margin of the ridge, of magnetic anomalies which are fairly shallow and are Donnelly13 suggested that the eastern margin of the Vene- parallel to the boundary. These probably represent the mag- zuelan Basin is a zone at which basinal crust had been netic expression of high-angle faults which are parallel to subducted beneath the Aves Ridge during the Cretaceous the boundary. The supra-crust sediment thickens dramati- and Paleogene. cally approaching the ridge51. For these reasons, and be- One of the most important attributes of the Venezuelan cause there are Cretaceous-Paleogene rocks of apparent Basin is the presence of a widespread seismic reflector

42 THOMAS W. DONNELLY

called horizon B". This reflector occurs between 0.5 to 1.5 vador, northern Nicaragua; see Burkart, chapter 15, this km depth, which is appropriate for normal ocean crust. The volume), which have a presumed Precambrian/Palaeozoic horizon is also a very efficient acoustic reflector, which is basement overlain by more or less thick Cretaceous and typical for basalt beneath sediment. However, the reflector early Cenozoic sedimentary rocks. This crustal material is appears to be far smoother in profile than typical oceanic submerged over much of the northern part of the Nicaraguan crust. Horizon B" was found during deep-sea drilling (Leg Rise. The dominant topographic features of the northern 15) to correspond with the upper surface of what was Nicaraguan Rise are vast, shallow banks of young carbonate subsequently identified as a plateau basalt. sediment evidently lying on continental crust. These banks The Venezuelan Basin shallows to the west; its western are essentially unexplored, except for some exploratory, boundary is called the Beata Ridge, which is a highly petroleum holes on the western margin that have encoun- 32 asymmetric topographic feature with a steep western face tered Eocene conglomerates and sandstones . The nature and a very shallow eastern face. The western boundary is of the buried boundary zone between the Nicaraguan Rise not a simple submerged cliff edge. Instead, there are several and the island of Jamaica remains totally enigmatic. horsts and grabens parallel to the western edge, but located The deeper portion of the Nicaraguan Rise is an up- within 100 km of the ridge itself. stepped extension of the Colombian Basin across the Hess Escarpment. A continuation of the B" horizon (defined for Beata Ridge the Venezuelan Basin) west of this scarp was demonstrated The Beata Ridge represents the western edge of the by drilling at DSDP Site 152, in the eastern portion of the uptilted Venezuelan Basin and is somewhat accentuated by rise. The nature of the boundary between the clearly oceanic horsts and grabens running parallel to the western escarp- and plausibly continental portions of the rise is completely ment. The ridge is the western boundary of a basalt escarp- unknown. ment which shallows gradually to the faulted edge. The face of the ridge must contain the outcrop of horizon B", but this Cayman Trough has not been directly observed. Fox and Heezen27 reported The Cayman Trough is the deepest (at least two regions the dredging of basaltic rock at depths of 2500 to 3200 m on are deeper than 6 km) part of the entire Caribbean. It is the western wall of the ridge; this must represent sub-B" bounded on the north and south sides by steep walls, which basalt. This discovery parallels the recovery of basalt at have the morphology of steep fault scarps, with a strong DSDP Site 15120, on the crest of the ridge, at a depth of 2400 overall resemblance to a transform fault. However, all other m below sea level. The western face has slopes of 8° to 15°. known transform faults are interruptions of the mid-ocean ridge system, and no traces of such a ridge can be found on Colombian Basin either side of the Cayman Trough. However, the pattern of The Colombian Basin is narrower and topographically earthquake distribution and the first motions of the stronger more complex than the Venezuelan Basin. On the east its earthquakes reinforce the view that this trough does indeed boundary with the Beata Ridge is sharp and has the appear- have a transform origin. The seismicity extends along the ance of a normal-faulted escarpment. On the west it is northern wall westward from Cuba to the middle of the separated from the lower Nicaraguan Rise by the low Hess trough, then steps across the trough along anorth-south axial Escarpment. There is no apparent difference in crustal type zone, and then continues westward on the south side of the between the Colombian Basin and the Nicaraguan Rise trough. The first motions of the quakes along the northern based on seismic information, and the reason for the fault and southern walls are right lateral, which is appropriate for boundary separating the two domains is not understood. a transform origin for this narrow basin. The Colombian Basin is occupied in the centre and The bathymetry within the deep Cayman Trough is northern part by a wide and evidently thick turbidite plain chaotic, apparently consisting of poorly defined north-south which has hampered geophysical techniques and which ridges, which are most conspicuous in the central (axial) part would prevent drilling. On the southwestern end the basin of the trough. The seismic axial zone corresponds with the terminates against a somewhat shallower (water depths shal- shallowest (about 3.5 km) depths in the trough, deepening lower than 3 km) deformed belt on the Caribbean side of Panama and Costa Rica.

Nicaraguan Rise Figure 3.2. (Next page) Map of the Caribbean The shallow portions of the Nicaraguan Rise appear to showing depth contours in km. The lowest con- be the offshore extensions of the continental structure of the tours of closed basins are hachured. Turbidite Chortis Block (Honduras, southeastern Guatemala, El Sal- plains are shown by a stipple pattern.

43 The Caribbean Sea Floor

44 THOMAS W. DONNELLY

towards the flanks of the axial high. There are also some was the site for the development and testing of many geo- ridges and secondary troughs more or less parallel to the physical techniques. Thus, many of the geophysical surveys north and south walls of the Cayman Trough; these evi- within the Caribbean were carried out at very early stages dently represent fault slices parallel to the major bounding of their technical development, and the history of Caribbean strike-slip faults of the trough margin. marine geophysics is very nearly a history of marine geophysics itself. These early studies had the advantage of Yucatan Basin bringing to the geological world knowledge of the Carib- The Yucatan Basin is smaller than the Venezuelan or bean crust at a very early time. The attendant disadvantages Colombian Basins. It is slightly shallower in the southern were that this information could not be placed in a world- half, where it merges with a poorly defined ridge on the wide context until surveys of other areas were made, and north side of the Cayman Trough (Cayman Ridge). In the that techniques were commonly improved after their early northern half of the basin, and mainly in the western part, it Caribbean trials, but the improved techniques were infre- is floored with a large turbidite plain. The northern margin quently applied at a later time to the Caribbean itself with Cuba is of indeterminate origin in most places, but is probably convergent in the northeast55. The margin on the Bathymetry western side, against Belize and Mexico, is steep and highly The earliest, and in many ways the most important, faulted. It has the appearance of a transform margin, possi- technique was simply the exercise of surveying the depths bly related to the opening of the basin in a northnortheast- of water of the oceans ('bathymetry'). On land this is a southsouthwest direction and the northern movement of simple exercise, and surveyors began producing numerous Cuba during the early Cenozoic. detailed topographic maps of the land surface in the mid 1800s. The same exercise is very difficult in the oceans, where man cannot venture far beneath the surface, and GEOPHYSICAL INVESTIGATIONS OF THE where light cannot penetrate more than about 100 m in any CARIBBEAN CRUST case. The invention of the sonic depth finder in the 1920s introduced a method which was not immediately used for Introduction general surveys, but which produced detailed depth profiles Water-covered portions of the Earth are not accessible along the tracks followed by ships, which tend to follow the to the sort of observations with which ‘classical’ geologists same shortest-distance tracks between ports. It was not until are familiar. The difficulty of making observations on an the late 1930s that general surveys were attempted. The ocean floor itself (which averages nearly 4 km in depth) is Caribbean was the site of one of the first surveys, and the actually the smallest part of the problem. The more serious contour map which was published in 1939 (U.S. Naval part of the predicament is that the oceans are the site of Oceanographic Chart of the Caribbean) was one of the first continual sedimentation almost entirely unrelieved by epi- such maps in the entire world. The difficulties of persuading sodes of erosion. Whereas continental and island areas are ship's captains to follow circuitous paths for surveying being continually subjected to erosive processes which strip purposes were overcome in the Caribbean in part by Franklin off the younger cover and exposes older rocks from which Roosevelt's love of deep sea fishing. It is said that several of geological deductions can be made, the ocean floor exposes his fishing trips were used also for surveying, because the fish virtually no rock or sediment which is not Holocene in age. were as good in one place as another , and he didn't care Our understanding of oceanic geology only began in where the ship went! very recent years, when 'remote sensing' techniques became available for detecting and defining the physical prop- Gravity erties of these doubly hidden older rocks and sediments. Other early geophysical efforts in the Caribbean in- These techniques, which mainly use acoustic sensing, are cluded submarine gravity surveys using a technique devised referred to as ‘geophysical surveying’. So successful have by F.A. Vening Meinesz of the Netherlands. Until the late they been that they have virtually created a new field of 1950s, submarines had to be used for marine gravity surveys geology in the last few decades: marine geology. The infor- because the motion of the waves made surface ships unsuit- mation that they provided and the questions that they raised able for a technique that requires a very quiet platform for were followed in 1968 by deep-sea drilling, which has the gravity meter. The major result of the surveys in the enabled geologists to obtain cores of sediment and igneous vicinity of the Caribbean was to show that the east-west rock beneath the ocean floor, and thus to provide tangible trending gravity minimum of the Puerto Rico Trench, which proof of geophysical inferences. has a large mass deficiency because the great depth of sea Because of its location and calm waters, the Caribbean water (which is less than half the density of oceanic sedi-

45 The Caribbean Sea Floor

Figure 3.3. Map of the Caribbean crust showing thickness of the solid crust Contour interval 5 km. This is based mainly on the results of seismic refraction surveying. Redrawn from Case et al.7. ment and about a third that of crust), bends south and transmitted by radio and recorded on the same record on parallels the Lesser Antilles island arc. This mass deficiency which the water transmitted pulses are recorded. Thus the is now explained as resulting from subduction of the Atlan- records show travel times of sonic phases, each of which has tic crust beneath the Lesser Antilles. The subduction process travelled through a different stratified acoustic medium (the drags (at the rate of only a few cm per year) a mass of shallowest layer is the ocean itself). If the media are of sediment westward and downward beneath the Lesser An- constant acoustic velocity and stratified such that velocity tilles. Compression of this sediment thickens and deforms increases step-wise in a downward direction, a successful the sediment mass, and part of the deformed sediment pile survey will enable geophysicists to calculate the depth of the has been raised above the ocean as the island of Barbados. top of each sediment or rock layer at the ends of the profile, The surveyed axis of maximum mass deficiency passes very as well as its seismic velocity. The two-ship seismic refrac- close to this island. tion technique was highly useful in obtaining gross crustal velocities and finding the depth to the MOHO (the acoustic Seismic Refraction horizon which separates the earth's crust from the mantle). The earliest geophysical technique which yielded The technique is much less useful for obtaining variations geologically useful information within the Caribbean itself in velocity within individual layers, and it is only marginally was two-ship seismic refraction. Following the develop- useful in obtaining velocities of the shallow pelagic sediment of the seismic refraction technique in the western ment cover. Atlantic in the later 1940s, mar ine geophysicists at Lament This method was widely used from the late 1940s Geological Observatory and Woods Hole Oceanographic (when large quantities of surplus World War II explosives Institution turned their attention to the Caribbean and Gulf were available at no cost) until the 1960s. The improvement of Mexico. In the two-ship technique one ship remained of acoustic systems which could detect much smaller signals stationary in the water with a sensitive hydrophone deployed (an air -gun using a large air compressor is used presently), and recorded the arrivals of sound pulses generated by the plus geometrically elaborate hydrophone arrays which other as it steamed away (over distances of several hundred could be deployed from a single ship, have made the two- km) setting off large explosions of TNT in the water. At the ship method obsolete, but its results had a profound effect end of its run the 'shooting' ship stopped dead in the water on the course of study of the ocean basins in general and in and deployed a similar hydrophone, becoming the recording the Caribbean in particular. ship as the first ship steamed towards the first, and became the shooting ship. For each shot the time of explosion is

46 THOMAS W. DONNELLY

Seismic Reflection the reflected sound took longer to arrive. The sound re- Another technique, which has become the most com- flected from the sea floor and sub-bottom reflecting hori- monly employed, is seismic reflection. This technique is zons took increasingly longer to reach the hydrophone, and similar to refraction except that the shot point and sound the trace of the reflectors on the recorder resembled hyper- detector are located nearly side by side. Thus, any sound bolas. As the distance increased beyond a certain point, received is reflected and not refracted. The reflection tech- reflected sound was no longer detected, because of critical nique was developed mainly during the 1950s, although the reflection (which is analogous with a familiar optical phe- first sub-bottom acoustic reflections had been recorded be- nomenon), and the received sound was entirely refracted. fore World War II. The earlier applications were mainly in The utility of this technique is that it yields sediment layer shallow water and used a variety of sound sources31. The velocities during a normal reflection profile. This technique application of the seismic profiling method to deep water26 was first known under the name 'sonobuoy' (which is the can be said to have been born in the Caribbean: the earliest name for a disposable floating hydrophone-radio combina- published deep-sea seismic reflection profiles in the world tion). The sonobuoy technique is especially valuable for are in the Yucatan Basin and Gulf of Mexico. obtaining acoustic velocities of the supra-crustal sediment, The seismic -reflection technique became possible but it often reveals the acoustic velocity of the uppermost when repetitive sound sources with sufficient energy to basement layers. penetrate the sea floor became available in the 1960s. The The sonobuoy method has evolved more recently into earliest sound sources were electrical sparks (which have a number of specialized techniques using arrays of hydro- low energy) and half-pound blocks of TNT, which are phones towed by the ship at varying distances from the hazardous. The present sound source of choice is the air gun, sound source. The results of these new techniques are often which emits a pulse of sound every ten seconds or so while described under the general and not highly descriptive term profiling. The returned sound reflections are sensed by a of 'seismic stratigraphy'. As detection devices (mainly to linear array (called a 'string') of receivers (hydrophones) sort out faint sound signals from a high ambient noise level) whose relative position strengthens the sound signal coming and computational techniques have improved, the new tech- from directly beneath the array and minimizes sound trav- niques now yield more information than was obtained dec - elling from other directions (such as the ship itself). The ades ago with expensive and dangerous two-ship seismic received sound is filtered to remove unwanted frequencies, refraction. particularly those resulting from the noise of the ship mov- The interpretation of sedimentary layering from seis- ing through the water. Recent applications of computer mic reflection is a relatively new and rapidly evolving processing have resulted in the acquisition of seismic profile geological exercise. Sound reflections in layered sediments records of exceptional detail, and the rate of improvement (or igneous rocks) result from contrasts in acoustic imped- of the technique is still rising. Modern computing tech- ance of the layers. Basically, acoustic impedance is the niques are capable of displaying not only an image of the product of the density and acoustic velocity of the materials. acoustically distinguishable layers beneath the sea floor, but In most cases acoustic impedance contrasts results from can also provide seismic velocity and other sorts of acoustic lithologic changes and thus record the stratigraphic layering information about individual layers. directly. These changes may be dramatic (a basalt layer Whereas the objective of seismic refraction is determin- beneath pelagic ooze; a turbidite layer within pelagic ooze) ing the acoustic velocities and thicknesses of the relatively or they may be more subtle (varying carbonate content in a deep crustal layers, the seismic -reflection technique is most pelagic ooze). The earlier surveys gave relatively crude useful in depicting intricacies of the layering of the overly- results and identified only a few 'horizons', as the reflecting ing pelagic sediment cover. The earliest seismic profiling interfaces were known. More recent surveys display com- had as its objective only the imaging of sedimentary layer- plex layering with detailed velocity information. ing. However, the technique quickly evolved into a hybrid refraction/reflection method capable of measuring acoustic Magnetic Surveying velocities, especially of the shallower layers. In this later The method of magnetic surveying is an additional version of the technique, a hydrophone and radio for trans - technique which was pivotal in oceanic research in unrav- mission of the received sound was attached to a float and eling the complexities of sea-floor spreading, but which has left stationary in the water as the ship steamed away and yielded very little useful information on the Caribbean crust. continuing to emit sound pulses with an air gun. When the This technique employs a sensor which detects differing float was close to the sound source (that is, at the beginning intensities of the ambient magnetic field ('magnetometer'). of the run), the received sound was purely reflected. As the This sensor has to be towed far behind a ship, because the distance between the sound source and receiver increased, magnetic field of the ship would interfere with the survey.

47 The Caribbean Sea Floor

Figure 3.4. Map of the Caribbean showing small topographic lineaments (solid lines) and small scarps (solid lines with hachure marks on the downthrown side) mapped on the sea floor. This map is adapted from Case and Holcombe6. The location of the earthquake studied by Kafka and Weidener37 is shown by the letter 'E'. The lineaments are interpreted as neotectonic responses to the continuing internal deformation of the Caribbean plate. In the Venezuelan Basin they parallel the magnetic anomaly pattern (not shown), and the isopachs of sediments of the B"-A" interval as well as the post-A" interval19. The existence of recognizable lineaments on the turbidite plains emphasizes the young age of these small tectonic features.

Alternatively, the sensor can be mounted in or behind an Venezuelan, Colombian and Yucatan Basins airplane. (Few people remember that the famous Reykjanes The Early Years Ridge survey in the early 1960s was made with an air-borne Refraction surveys in the Caribbean have provided the magnetometer flown a few hundred feet above the ocean most important geophysical evidence bearing on the prob- surface). The magnetic intensity in the main ocean was lem of the origin and nature of the Caribbean crust. It is found to consist of the ambient Earth's field with either a difficult to realize that all of the published surveys were small addition or subtraction caused by oceanic crust per- undertaken in the interval of only a few years. (Two earlier manently magnetized either parallel or anti-parallel to the refraction lines by Sutton in the Venezuelan Basin have earth's field axis (a so-called reversal). The spatial pattern evidently not been published, in spite of references suggesting of these reversals was found to correlate with the temporal the contrary.) Ironically, the first published Caribbean history of reversals of the Earth's field, which was the major seismic refraction line (executed in 1953), which was in the conceptual hurdle to deciphering the story of sea-floor Yucatan Basin, remains the only line in this basin to the spreading. present date! In 1955 there were separate surveys in the Venezuelan50 and Colombian Basins, Nicaraguan Rise, and Cayman Trough23,24. In 1956 there was a further survey in REGIONAL GEOPHYSICAL CHARACTER OF 51 THE CARIBBEAN the Venezuelan Basin , and in 1957 another survey along the South American borderland and across the Beata Ridge The following account summarizes knowledge of the Car - and eastern Colombian Basin18. These were the last two-ship ibbean crust and pelagic sediment cover made initially seismic refraction profiles obtained in this area, but the density through seismic refraction and reflection techniques, which of refraction information remains one of the highest for the was later refined with the results of deep-sea drilling. Be- entire world ocean. cause I will concentrate on the Caribbean crust itself, I will Edgar et a/.18 presented the final results of this tech- omit several excellent papers covering the borderlands40,60, as nique and summarized the state of knowledge (mainly thick- well as most of the material in Biju Duval et al.2 ,Ladd et ness of crust) obtained from these methods. The most al.44 and Ladd and Watkins42. startling revelation is that the crust of the Caribbean is

48 THOMAS W. DONNELLY

oceanic in character, but much thicker than 'normal' oceanic Ewing's horizon A" was found to be Middle Eocene cherts crust (Fig. 3.3). Whereas normal oceanic crust has a thick- at the top of a sedimentary sequence rich in radiolaria. ness of about 5-6 km (which, along with 0.5 to 1 km of Horizon B" was identified on Leg 15 as the top of a wide- sediment cover, means a depth to the MOHO (lower bound- spread basalt sequence of late Turonian (in the east) to early ary of the crust) of about 6-7 km), in the Caribbean this depth Campanian (west) age. The main aim of Caribbean marine was commonly greater than 10 km and in the Beata Ridge geophysical investigations since Leg 15 has been to inves- areaabout20kmdeep. Officer et a/.50 speculated that "...the tigate the implications of the B" horizon, which was recog- entire Caribbean area may have been extensively intruded nized immediately as not representing normal oceanic crust. by large bodies of primary basalt magma...". This was one of the most prescient and significant discoveries in the Modern Geophysical Surveying history of Caribbean geology. Subsequent marine geophysical investigations have re- Ewing et a/.25 summarized several years of early seis- flected changing techniques in seismic surveying. There mic reflection profiling in the Caribbean and recognized two have been three notable sonobuoy studies in the Caribbean. 46 prominent reflection horizons. They named these the A" and Ludwig et al. presented the results of a 1970 cruise in the B" horizons; the identification and elucidation of thes e have Colombian and Venezuelan Basins, showing that the sedi- subsequently been pivotal in Caribbean marine geological ment and crustal acoustic velocities along this east-west line and geophysical research. The A" horizon was a persistent matched the velocities of the drilled materials recovered at and widespread acoustic reflector which was located at DSDP Sites 146 and 151 (Leg 15). Later cruises in the 36 62 about half the total thickness of the sediment layer, and the Colombian Basin and southern Venezuelan Basin em- B" horizon was the acoustic basement. After the initial phasized seismic reflection, but included extensive description of these reflectors, their explanation became a sonobuoy observations. These cruises showed that the major problem of Caribbean marine geology. The A" reflec- acoustic velocities of the Caribbean crust and supra-crustal tor obviously represented an important lithological break sediments were more variable than suggested by the results 46 between two contrasting sediment types, but several plausi- of Ludwig et al. ble sedimentary contrasts might yield similar reflectors, and There also exists an extensive body of recent studies the acoustic record could not choose among the possibilities. which have employed seismic reflection profiling to image The B" horizon resembled 'normal' oceanic crust in that it the sediment layering, and from which important geological is several hundred metres below the sea floor and is a highly inferences could be drawn through construction of isopach efficient reflector, letting very little sound through. How- maps and similar techniques. Edgar et al19 presented a ever, it was far smoother in profile than normal oceanic profiler study of the Venezuelan Basin, for which a more crust, which has a characteristic hummocky appearance. complete data set was published by Matthews and Hoi- Almost from its first appearance on seismic profiles it was combe47. Recognition of the A" and B" horizons enabled considered to be something other than normal oceanic crust. the thicknesses of the post-A" and A"-B" intervals to be determined rapidly, so that they could be plotted along the Deep Sea Drilling Project lines of the seismic profiles. These profile records showed The development of the seismic profiling technique that the thickness of these two sediment layers changed came fortunately only a few years before the initiation of the abruptly along the survey lines. Deep Sea Drilling Project, which is arguably the single most Variable thicknesses of these layers can be interpreted important geological research project in this century. One in several ways. First, there might have been an original of the first aims of the project was the identification through difference in sedimentation rate (rate of production of bio- recovery by drilling of the lithologies which corresponded logical sediment; rate of accumulation of fine-grained clas- to the reflection horizons found in seismic reflection profil- tic sediment from the land areas). However, these ing. In the Caribbean there were two drilling legs (Leg 4, differences are not expected to be abrupt along a given line. February-March, 1969; Leg 15, November 1970-January A second possibility is that the differences in sediment layer 1971) devoted in part (Leg 4) and completely (Leg 15) to thickness record subsequent erosion in response to bottom the Caribbean (Fig. 3.la). A later site (DSDP Site 502, Leg currents. These are referred to as hiatuses in sedimentation. 68, August 1979), in the Colombian Basin, was drilled in Evidence that this has been the case is found in drilled part as a follow-up of a hole drilled during Leg 15 and also sediment sections in the southern Venezuelan Basin (espe- as the refinement of a new drilling technique. I will discuss cially site 150, Leg 15, which found several stratigraphic only the five holes (all on Leg 15) that penetrated the hiatuses). These, however, are not expected to be abrupt sediment cover and reached the basement of the Caribbean along a line. Third, vertical faulting during sedimentation crust. might result in erosion of the sediment cover over the up- One of the first discoveries of drilling on Leg 4 was that thrown fault block, as a result of sediment erosion caused

49 The Caribbean Sea Floor

Figure 3.5. Cartoon sketch of east-west cross section of the Caribbean showing the inferred cross-section of the Caribbean Cretaceous basalt province, based largely on the information of Figure 3.3. The lower boundary of this plateau basalt merges imperceptibly with the normal oceanic crust. The change of thickness in an east-west direction is inferred from seismic refraction measurements of the depth to the MOHO (mantle). The small triangular features on the surface of the basalt plateau in the east are volcano-like mounds. by increased bottom-water velocities over the up-thrown of the sonobuoy stations of Talwani et al.62, thus allowing block. The change of thickness in a sediment layer might be a useful comparison between techniques. Ladd and Wat- quite abrupt across the line of the fault. Evidence that this kins 43 presented some velocities obtained by seismic strati- had been the case was shown by Edgar et al19 , who dem- graphic techniques. Lu and McMillen45 added some onstrated that sediment thicknesses (represented as velocities in the western Colombian Basin. The implication isopachs, that is, contours of equal thickness for a given of these most recent surveys is that the lithologies of the acoustically-defined sediment layer) thinned abruptly in sediments are more heterogeneous geographically (as well either a northeast-southwest or a northwest-southeast direc- as stratigraphically) than had been presumed. Until further tion. Minor topographic lineaments of these orientation can deep-sea drilling is undertaken we will not understand the be seen in the present sea floor (Fig. 3.4), which evidently meaning of these heterogeneities. indicates that this faulting continues to the present time. Ladd et al.44 and Ladd and Watkins42 presented pro- More recent sedimentary profiling observations have files of the northern and southern boundaries of the Vene- focused mainly on seismic stratigraphy techniques, and zuelan Basin which showed that the floor was being include the results of Ladd and Watkins43 in the Venezuelan subducted beneath the Greater Antilles and South America, Basin; Diebold et al.10 in the southeastern Venezuelan Basin respectively, a conclusion also reached by Talwani et al62. (which employed an elaborate two-ship 'expanded profile’ Biju Duval et al2 presented profiles of the Venezuelan method somewhat similar to the old two-ship refraction Basin, as well as the Aves Ridge and Grenada Basin. Al- technique, but in which the ships cruise exactly away from though one could conclude that there has been eastward-dip- each other simultaneously); Stoffa et al.61 in the western ping subduction of the Venezuelan Basin floor beneath the Venezuelan Basin; and Lu and McMillen45 in the southern Aves Ridge, the seismic evidence for this is not clear. The Colombian Basin. Diebold et a/.10 resurveyed sites of four Grenada Basin, which is a deep basin between the Lesser

50 THOMAS W. DONNELLY

Antilles and the Aves Ridge in the southeast comer of the gesting either a somewhat abnormal crustal section or the Caribbean, remains enigmatic and in need of modern sur- vagaries of dredging in a topographically restricted and veying techniques. It may be a 'pull apart' basin, as sug- complex place. If the crustal layers were horizontal and gested by its thick sediment fill. uncomplicated, it would be concluded that the diving results Bowland and Rosencrantz4 presented profiles of the show a very thin basalt layer above gabbro. However, the western Colombian Basin with rock units identified, but, in recovered rocks probably are from faulted slivers and do not the absence of velocity or similar information, the identifi- represent simple horizontal layering. While the Holocene cations are problematic. Similarly, Rosencrantz55 presented age and Neogene opening of the trough has not been in detailed profiles of the Yucatan Basin, again lacking velocity serious doubt since the original exposition by Holcombe et information. Because these latter areas have not been a/.34, the total history of opening remains undetermined. drilled, even the approximate ages and lithologies of the The central zone, extending about 150 km either side of layers are unknown, and the age and nature of the basement an axial high (for a total width of about a third of the entire is elusive. The thickness of supra-basement sediment is trough) deepens exponentially on either side of the axial much thicker in the Colombian Basin than in the Venezuelan high. Holcombe et al.34 fitted thermal subsidence models to Basin. Part of the explanation probably lies in greater depo- the bathymetry and concluded that the best fit of these two sition of pelagic carbonate-rich sediment during the late curves implies a total spreading rate of 2.2 cm yr-1. A Cretaceous and early Tertiary, but an additional part of this problem is that the depth at the mid-trough ridge is about thickness might be clastic debris shed from Central Ameri- 3.5 km, which is about 1 km deeper than normal mid-ocean can as it rose during the late Miocene to Recent. ridges. The narrowness of the Cayman Trough implies that the heat loss must be accommodated in lateral as well as Cayman Trough vertical dissipation, allowing a probably higher loss rate and The Cayman Trough is a special portion of the Carib- higher subsidence (that is, greater water depth over the ridge bean crust which has been the focus of studies aimed at an axis). However, the applicability of a modified mid-ocean understanding of this very deep, linear trough. Bowin3 model is necessarily somewhat in doubt under these extreme summarized the results of his gravity observations with the conditions. 56 earlier seismic refraction results of Ewing et al.23 to produce Rosencrantz et al. have presented the most refined a structural model of the trough, finding, among other model for the opening of the Cayman Trough to date. They things, that a thin crust and correspondingly shallow MOHO propose four ridge positions since marine magnetic anomaly required by isostatic compensation is indeed found in the 20 (middle Eocene) and three ridge jumps to the present axis of this trough. Heat flow values22 were found to be high position. The magnetic data is much less convincing than for compared with average world-wide oceanic values (which most other ridge systems, including several large anoma- average 1.6 x 10-6 cal cm-2 s-1), suggesting young oceanic lies that are disregarded in their analysis. Heat flow values crust. A major breakthrough came with a paper by Hol- along the trough do not show the expected exponential combe et al.34, who showed that the axial topography, and decay trends. While the overall conclusion of Neogene especially the exponential deepening away from a shallow spreading transverse to a spreading axis is convincing, axial zone, led to the conclusion that this trough was the site largely on purely bathymetric grounds, further inferences of modem sea-floor spreading. Holcombe and Sharman33 about earlier spreading histories, with several ridge jumps, and Rosencrantz and Sclater57 refined these observations are not so convincing. The age of opening of the trough and and showed a complex spreading pattern. the history of opening rates remains enigmatic. Geological There is little doubt that the Cayman Trough represents problems with a 1000 km opening during the Eocene and more or less normal, but very thin, oceanic crust formed at extending westward to Guatemala have been discussed by 13,15 an abnormally great water depth in a narrow basin. Dredged Donnelly , who suggested that much of the opening rocks 8 include a disproportionate fraction of gabbros, sug- might have taken place concomitant with internal spreading

Figure 3.6. (Next page) Magnetic total intensity profiles across features of the Caribbean. (A) DSDP Site 145, northern Venezuelan Basin (from Raff54). (B) Air-borne magnetic profile of the southern peninsula of Haiti showing magnetic signature of the Dumisseau Formation. Computed and plotted from original project MAGNET records. Both profiles show the distinctive pattern of remnant magnetization with a nearly horizontal vector, which is characteristic of the mid-late Cretaceous magmatic bodies in the Caribbean.

The Caribbean Sea Floor

52 THOMAS W. DONNELLY

of the Chortis block. tism at the eastern sites ended contemporaneously, but continued for an additional few million years in the west. CHARACTER OF THE CARIBBEAN CRUST Ludwig et al.46, whose sonobuoy survey line passed close to drilled site 146 (coincidentally their sonobuoy As recently as 1954, the Caribbean was considered by some to station 146 was closest to this drill site), found that the total be continental crust converted to oceanic in relatively crustal thickness was shallower in the east and not very young geological time65. Ironically the first seismic refract- different from normal ocean crust. Conversely, workers in the tion observations in the Caribbean had already been made; western Venezuelan Basin and Colombian Basin36,61 have these showed definitively that the crust of the Caribbean was continued to point out the great thickness of the crust there. It undoubtedly oceanic 23,30,51. However, these same surveys is now appreciated (Fig. 3.5) from a variety of observations also showed that the oceanic crust was abnormally thick for that the excessive thickness of the Caribbean crust diminishes much of the Caribbean, an observation that was not fully to the east. appreciated until later drilling. The smoothness of the B" horizon in profile had been One of the most important results of DSDP Leg 15 was emphasized in the earlier papers. Later refinements (higher the discovery of late Cretaceous basalt in five drill holes in source energy, improved signal to noise ratio) in the seismic the central Caribbean20. This material was quickly identified profiling technique have revealed a rougher texture than was to be of the same age and character as numerous basalt formerly apparent. Relatively modern surveys have found occurrences on land areas surrounding the Caribbean, and small reflection hyperbolas on B", showing some topography the Caribbean was identified as the site of widespread on the layer. (Unfortunately, Talwani et al.62 penned these in ‘flood’ basalt eruptions11,17. heavily on their profiles, thus robbing their readers of the opportunity of making this judgment for themselves.) Geophysical character of the Caribbean crust Nevertheless, the contrast between fairly smooth B" and Ewing et al25 designated the deeper of two seismic typical, rough, oceanic crust remains apparent throughout the reflection horizons in the Venezuelan Basin as B", which surveyed areas. From this it is concluded that the basalt was they compared to the previously-named horizons B in the erupted either in sheets which spread great distances ('flood Atlantic and B' in the Pacific. Horizon B" appeared to be far basalts') or which intruded the shallow pelagic sediments smoother than oceanic crust and it was not interpreted as this which covered the basin floor. The density of liquid basalt material. The material underlying this horizon was the crus- (around 2.9 g cm-3) is much higher than water -rich pelagic tal material found in earlier seismic refraction sur- sediment (density about 1.5 g cm-3, but variable). It would be veys23,50,51. The occurrence of the basalt drilled during Leg physically impossible for a dense liquid to rise through a 15 matched seismic refraction survey results18, in which a low-density, strengthless sediment layer. Thus, erupting crustal or quasi-crustal material with velocities greater than basalt in a basin floored with pelagic sediment would have to 5 km s-1 occurred at this level. Thus, the top of the abnor- be intrusive. mally thick Caribbean crust was identified. One interesting feature of the B" horizon is the occur- During DSDP Leg 15, horizon B" was penetrated in five rence of a few buried sea mounts in the northern Venezuelan drillholes (the deepest only 16 m into the basement) and the Basin. One of these was to have been drilled by the DSDP material which provided the large acoustic impedance con- (site 145), but the drilling was aborted because of mechanical trast for the seismic horizon was found to be basalt and problems. This site has the magnetic signature of basalt of the dolerite20. Unlike the normal basalt of the sea floor, this was age of the recovered B" material; that is, it shows a strong somewhat coarse-grained and not pillowed. Furthermore, permanent magnetization with a low-inclination 54 there was no iron-enriched sediment above the basalt, an (Cretaceous) vector (Fig. 3.6a). The magnetization profile is observation not fully appreciated at the time because so few similar to that seen in the on-land Dumisseau basalt of Haiti penetrations to the crust had been made by the end of 1970. (Fig. 3.6b), which has been interpreted as an obducted fragment 49 In two sites the basalt included limestone inclusions, and the of the Caribbean crust . Similar sea mounts are shown as interpretation made on board the ship was that the basalt had circular low -thickness isopachs of the sediment layer between 19 been emplaced as shallow sills in pelagic sediment. The age B" and A" in the maps of Edgar et al. and Matthews and 47 of the sediment immediately above the basalt was latest Holcombe . The clearest seismic reflection record of one of 43 Turonian in three localities (DSDP Sites 146,150,153), that these sea mounts is shown by Ladd and Watkins (their age or slightly younger at DSDP Site 151, at which recovery figure 4, line VB-3-NB). The size (about 15 km across) of had been poor, and early Campanian in the westernmost this mound and its morphology (height about 750 m, slopes DSDP Site (152). Thus, there was a suggestion that magma- about 5°) agree well with a volcano. The failure to drill DSDP Site 145 remains a sad memory.

53 The Caribbean Sea Floor

Figure 3.7. Graphs relating the carbonate content of drilled sediments in the B"-A" interval of three Caribbean DSDP Leg 15 sites 20. (A) (top) Site 146, showing a decline of carbonate upwards in the late Cretaceous, reaching zero at the lower boundary of the Cenozoic. (B) (bottom) Site 152, showing high carbonate contents throughout. (C) (opposite, top) Site 153, showing a thinned Cenozoic section with scattered high and low carbonate contents.

54 THOMAS W. DONNELLY

pretation of magnetic anomalies consistent with this age and In the Colombian Basin the identification of B", or suggested an age of opening as late Cretaceous to Paleogene. acoustic basement, has been more difficult. Instead of a very 55 smooth reflector, commonly the reflecting horizon is rough, Rosencrantz presented a tectonic model which would or highly faulted, or too deeply buried to register clearly on have produced magnetic anomalies at approximately right reflection profiles. One of the drill sites (DSDP Site 152) angles to those of Ha ll and Yeung, which emphasizes the penetrated the B" horizon in the foot of the eastern Nicara- difficulty of mapping the faint magnetic lineations in this guan Rise close to its boundary with the Colombian abyssal basin. plain. Unfortunately, this leaves virtually all of the remain- Magnetic anomalies and structural lineaments in the ing Colombian Basin and lower Nicaraguan Rise with no Venezuelan Basin proof that the basement reflector represents the same basalt The discovery of magnetic anomalies in the Venezue- horizon as was found in the Venezuelan Basin. However, 12 4 lan Basin led several workers to interpret these as sea-floor Bowland and Rosencrantz extended the designation of this 28 horizon throughout the western Colombian Basin. spreading anomalies. Ghosh et al. found a symmetry in Seismic refraction results show that the entire crustal the anomaly pattern and fitted them to a sea-floor spreading section is thicker in the Colombian Basin than in the Vene- sequence of Jurassic to early Cretaceous age. The symmetry zuelan Basin. Figure 3.5 shows that the crust of the entire of the anomaly lines and the excellence of the fit is not as Colombian Basin-Beata Ridge area (including the plateau apparent to many workers as it evidently is to these authors. Also, there are serious problems with their interpretation. basalt) exceeds 15 km, and in the western Colombian Basin -1 thickens to beyond 25 km at about the point where the The inferred spreading rate (about 1 cm yr ) is about half superincumbent sediment begins to thicken sharply (see Lu the inter-American divergence rate for this period and an and McMillen45, Fig. 7, for sediment thickness). addit ional spreading centre to make up this deficit has not The crust of the Yucatan Basin has not been identified. been identified. Also, at least one of the original anomalies Ewing et al.23 showed, in an end-to-end seismic refraction was inferred, on the basis of its similar magnitude at differ- profile, that it has the velocities of and a thickness only ent elevations (3 km elevation (air-borne) and sea surface) 55 to come from a source too deep to correspond to the Carib- slightly greater than typical oceanic crust. Rosencrantz 12 stated that the basement seismic reflection is smoother than bean crust . For example, the sea-floor spreading anoma- that of typical ocean floor and suggested that it might be lies across the mid-Atlantic Ridge are rarely visible in 3 km air-borne surveys of the same project, except for the strong similar to B" of the Venezuelan and Colombian Basins. Hall and Yeung29 interpreted heat flow values21,22 as consistent axial anomaly. with a late Cretaceous crust. They also presented an inter- An additional observation which causes difficulty for a

55 The Caribbean Sea Floor

simple sea-floor spreading explanation for the magnetic Sub-B" acoustic reflections and the physical character anomalies is the parallelism between the anomalies and the of the plateau basalt conjugate northeast-southwest and northwest-southeast Hopkins35 reported a faint seismic reflection beneath structural pattern seen in the topographic lineaments (Fig. B". Saunders et al 58, in the same volume, reported addi- 3.4) as well as the isopach maps mentioned above. This tional examples, but Stoffa et al.61 dismissed one of these as pattern is aneotectonic record of modern deformation of the a multiple internal reflection. However, Stoffa et a/.61, plate. Burke et al.5 likened this pattern to a 'Prandtl cell'; Ladd and Watkins43 and Biju Duval et al.2 have all reported that is, a plastic body deformed by being squeezed on three these reflections and there is little doubt that they exist. The sides and being allowed to extrude in the direction of the reported reflections all have a low dip (if they were parallel fourth side. Thus, the lineaments are a conjugate set of small to B" it is doubtful that they would have been noticed), but strike-slip faults within the deforming body. They suggested not in a consistent direction. Most of these reports come that the lineament pattern resulted from the Caribbean crust from the southwestern Venezuelan Basin and the dips are to 37 spreading eastward. Kafka and Weidner studied an earth- the south or to the north, with a minor eastward component. quake (MB 4.7, depth 18 km, 14th August, 1972) within the The depths of the reflectors are variable, ranging from very plate and found a first motion corresponding to sinistral close to B” to about 0.5 s (2.5 km at 5 km s"1) beneath B". movement on a northwest-southeast lineament. They said Although there must be a contrast in acoustic impedance in that this contradicted the sense of motion postulated by order to have a reflector, there is no clear lithological expla- 5 Burke et al , but stated that perhaps an earlier fault had been nation for this contrast. Buried sediments are possible, but reactivated by modern east-west compression and is now layered hyaloclastics are equally plausible, and a textual undergoing a re verse sense of motion. Figure 3.4 shows the change in layered basalt itself cannot be dismissed. proximity of the earthquake to an appropriate (northwest- Additional evidence for the character of the sub-B" may southeast) linear element of the surface lineament pattern. be inferred from the velocity structure of the entire supra- The surface lineament pattern is also parallel to sedi- MOHO section. Earlier two-ship seismic refraction veloci- ment isopach contours shown by both Edgar et al.19 and -1 47 ties showed velocities between 6 and 7 km s for deeper Matthews and Holcombe . Both of these papers referred to parts of this layer, but these velocities are themselves not the same data set, but the latter paper displays the isopachs very revealing. Diebold et al.10 have shown velocity profiles over a larger area. These isopach contours show that both that are not unlike oceanic crust, but with widely differing the pre-A" and post-A" sediment (that is, 85-50 Myr and thicknesses of layers. More intriguing are the lower (4-5 km s- 50*0 Myr ago). Thus, during the last 85 Myr, the Venezuelan 1) velocities found by Ludwig et al.46 in the shallower Basin has been riven by faults which apparently remain portion of the basement, especially in the central Venezue- active today. There is little doubt that the lineament pattern lan Basin. The depth of these layers places them squarely in results from deformation similar to the 'Prandtl cell' model 5 the depth range of sub-B" materials, but the velocities seem of Burke et al. . Because the grain of the magnetic anomaly too low to match the drilled materials recovered in Leg 15. pattern is parallel to these faults, I consider that these mag- They could be hyaloclastic basalt or they could represent netic anomalies most probably originate subsequent to the included sedimentary layers within the upper part of the formation of the crust and do not represent sea-floor spread- sub-B" layer. The normal oceanic crust has similarly low ing anomalies. velocities near the ridge crest where cementation has not yet The origin of the magnetic anomalies is still enigmatic. filled in the velocity-lowering fractures which are abundant Their apparent depth (based on a faulted slab or reversed in young basalt. Nowhere in oceanic crust of Cretaceous age polarity block model) still poses a major problem. The elsewhere in the world is there a clear counterpart of these magnitude of the anomalies (100 to 200 gammas) makes it velocities and the identification of this lower -velocity ma- difficult to find an appropriate faulted-slab model that will terial must await future drilling. explain them, considering that they would have to lie at a depth well below B". They may result from another phe- Lithological and petrological character of the plateau nomenon, with a different model signature. A possibility is basalt a dyke swarm orientated to one or the other of the fault Our knowledge of the sub-B" material comes not only directions of the conjugate set found in the Venezuelan from five shallow drill penetrations, but from extensive Basin. The notion that they are older crust (late Jurassic?) occurrences of correlative obducted ophiolite. One of the buried beneath a kilometre or more of late Cretaceous basalt unusual features of the Caribbean is the wides pread occur- seems unsupportable. rences of basaltic (=ophiolitic) complexes of Cretaceous age found on land in the general tectonic context of thrust sheets with more or less abundant serpentine. These are summa-

56 THOMAS W. DONNELLY rized in Donnelly et al.16. These have the appropriate age, absent from Phanerozoic terrains. Their appearance in lithology, and associated sedimentary rocks to make their the Caribbean indicates that the melting within the identification with sub-B" material inescapable. The identi- mantle acheived a very high temperature which implies fication of the sub-B" material in the Venezuelan Basin by a vast amount of excess heat at depth. its relationship with obducted ophiolitic material (which is It is safe to conclude, on the basis of seismic especially clear in Curasao, Aruba, Venezuela and His- reflection, drilled occurrences and obducted mafic paniola, and slightly less so in Puerto Rico and Trinidad) igneous complexes, that the entire Caribbean Cretaceous can be extended westward to land areas adjacent to the basalt province once occupied nearly the entire western basins, even though the westernmost drilled occur- Venezuelan Basin, Beata Ridge and Colombian Basin, rence of this material is far removed to the northeast (DSDP extending an unknown distance onto the Nicaraguan site 152 on the eastern lower Nicaraguan Rise). The finest Rise. A possible northern extension or equivalent has exposures in the west of this basaltic material are in Costa remains on Cuba and Guatemala Rica, but extensive outcrops can be found also in Panama, Colombia and Jamaica. Similar material, of similar age, is Conclusions regarding the plateau basalt found in Guatemala and Cuba; its identification with any of Donnelly11,13 concluded that the sub-B" layer of the the presently named Caribbean basins would be highly Caribbean represents a flood basalt province of vast original certain except for the fact that the Chortis Block (continental size (possibly about 1.2 km2), ranking it among the world's crust of Honduras and adjacent regions) now intervenes great Phanerozoic flood-basalt provinces. The age of termi- between these northern outcrops and the bulk of the southern nation of this event was about 85 Myr in the east and was occurrences. apparently younger, but not well-defined, in the west. In Prior to the drilling of Leg 15 it was not apparent what Costa Rica, where the western equivalents are best exposed these vast on-land basalt-serpentine masses signified. A on land, there seems to have been a severe decrease in major result of the drilling was to show the unity of the entire eruptive activity in the early Campanian, but minor magma- assemblage as a vast Caribbean flood basalt province11. The tism continued to the Paleogene. Nowhere on land is there a stratigraphic level of the supra-basalt sediments on the land clear indication of the beginning of the event. Late middle correlates with, as nearly as complex structural environ- Albian ammonites have been found within the basalt on ments allow, the stratigraphic level found for the termina- Curacao64, which remains one of the firmest dates within tion of the event in basinal occurrences. the eruptive dates itself. There are numerous finds of am- These on-land occurrences are dominantly pillow ba- monites, radiolarians and foraminifers16 in sedimentary salt, but include massive basalt, as well as locally thick rocks which are arguably related to some stage of the erup- hyaloclastic beds. The availablity of on-land examples has tive history for on-land occurrences, but in most cases some enabled the study of these rocks in far greater detail than if doubts can be raised that thes e do not place a firm age research was confined to the recovered materials in five drill constraint on the associated basalts. holes. An account of the petrology of this basalt complex One of the major elusive questions is the age at which was given in Donnelly et al.16 . In summary, this basalt is basaltic magmatism began in the Caribbean. The well-ex- thick (1 OOs of m or a few km) in on-land occurrences. Most posed and well-studied rocks of Costa Rica appear to proof the basalt has the chemical (major and minor elements) vide important evidence. In the Nicoya Complex of western attributes of normal, ridge-generated sea-floor basalt (mid- Costa Rica a radiolarian chert sequence appears to lie be- ocean ridge basalt or 'MORB'), but one of the drilled (DSDP neath the upper Nicoya complex (which is equivalent to the Site 151, Beata Ridge) and several on-land (Hispaniola, basinal basalt) and the lower Nicoya basaltic complex. Costa Rica) occurrences are of a basalt enriched in the However, even this excellently exposed sequence places incompatible elements, including the light rare earths. Near surprisingly few constraints on the basalt eruptive history, the base at one locality (Curasao) and possibly in the same because many, and perhaps all, of the contacts with the lower stratigraphic position at an isolated, insular site off the west complex are probably intrusive. The lower complex has coast of Colombia is a picritic basalt (called 'komatiite' at been considered to be older than the chert, whose earliest the latter locality) which has a high magnesium content age is Callovian or older. However, the lower complex is unparalleled among Phanerozoic basalts. Similar, but dominantly intrusive and could even be interpreted as a slightly less magnesian, basalts are found in Costa Rica coeval intrusive equivalent of the upper complex. Dozens This high-Mg basalt has a special significance. Basaltic of geologists labouring for several decades have not been melts of this composition require a high degree of melting able to solve the problem of the ages of the upper and lower and are erupted at particularly high temperatures. They were complex. erupted abundantly during the Archaen, but are virtually Thus, several major questions remain unanswered. What is the thickness and volume of erupted material? At

57 The Caribbean Sea Floor

what age did the enq>tion begin? What was the average and tropical Atlantic (DSDP Site 144 on the foot of the Demarara maximum rate of eruption (cubic kilometres per million Rise encountered carbonaceous pelagic sediments of late years, or some other measure)? Is there clearly defined older Turonian to Santonian age30). The unusual age and character crust beneath the basalt event elsewhere in the Caribbean? of these sediments appears to link the Caribbean (espe- It is of considerable interest, but little help, to note that cially Site 153) and Atlantic sites. To the extent that there are other, similar and possibly coeval basalt occur- carbonaceous sedimentation requires anoxic bottom waters rences elsewhere in the world. The Ontong- Java plateau of which in turn implies geographic restriction, these occur- the western Pacific has perhaps the highest similarity, but rences provide an obstacle to the view that the Caribbean there are others, also during the Cretaceous, which should basalt might occurred in the Pacific Ocean distant from a be considered. Evidently the basalt magmatic activity of the land mass. entire world was exceptionally high during this time period. The importance of the B"-A" interval is largely for the information it reveals on the mode of formation of the sub-B" layer, the Caribbean Cretaceous basalt event. We SUPRA-CRUSTAL PELAGIC SEDIMENTS have two sorts of information; the drilling results of DSDP Leg 15 and seismic reflection studies, including sonobuoy. Virtually all that is known of the pelagic sediments of the Leg 15 drilling found a 350 m-thick section between B" and Caribbean comes from drilling DSDP Legs 41, 1520 and 53 A" at DSDP Sites 146 in the Venezuelan Basin and 152 in 68 (which sampled only sediment down to Upper Mio- the lower Nicaraguan Rise. The correlative section at the cene). Complete descriptions of these sediments may be more southerly DSDP Site 153 (southwest Venezuelan Ba- found in these volumes. In only four sites (DSDP Sites 30 sin) was only 200 m thick. The remaining two penetrations and 148 on the Aves Ridge, and 154 and 570 in the western (DSDP Sites 151 on the Beata Ridge and 150, south-central Colombian Basin) are there, appreciable thicknesses of Venezuelan Basin) found highly thinned sections, with hemipelagic sediment deposited close to a continental area thicknesses of about 30 and 35 m, respectively. In the remainder of the sites the sediments are typical pelagic One of the most important questions for the early Car- sediments similar to those encountered far distant from ibbean is the determination of palaeodepth of the water A continental areas. There are only a few good occurrences of normal ridge-generated oceanic crust subsides as the square on-land sediments arguably correlative with these pelagic 49 59 root of time following crustal generation at the ridge axis. sediments (Haiti , Costa Rica ), compared with the larger This relationship (which explains most of the bathymetry of number of occurrences of the underlying basalt. normal oceanic crust) is called the "Sclater-Francheteau The persistence of a prominent acoustic reflector (A") curve" and is explained as resulting from thermal contrac - in the Venezuelan Basin enables us to divide the sedimen- tion resulting from the dissipation of excess (magmatic) heat tary section into a post-A" upper part and B"-A" lower part. brought to the ridge axis during sea-floor spreading. The The pelagic sedimentary section is highly diverse beneath Caribbean evidently formed not as a normal ridge, but as a reflector A", and more homogeneous above this level. vast basalt plateau. It is theorized that the water was shal- Sediments of the B" -A " interval lowest over the approximate centre of the province, where the basalt was thickest. However, the entire province should As noted above, B" is latest Turonian, or about 85 have subsided with time, as the original heat was dissipated. million years old, in the Venezuelan Basin and somewhat The thickness of the basalt and the time interval over younger in the Colombian Basin. Horizon A" is slightly time which it was erupted are unknown. However, the history of transgressive, being middle Eocene (about 50 Myr) in the thermal subsidence of the body can place some constraints east, but late Paleocene (about 58 Myr) in the west. on these values. As crust subsides it reaches a water depth The pelagic sediment sequence in this interval is domi- below which sinking calcareous debris (biogenic fossils) are nantly calcareous, but with a strong siliceous component. dissolved. This horizon, the CCD (=Car bonate Compensa- This interval shows great vertical thickness and variability, tion Depth), varies from place to place and time to time, but suggesting deposition within a topographically rugged area in a relatively small part of the world ocean and in a brief Occurrences of intercalated radiolarian turbidites and vol- interval it can be taken as a flat plane. Calcareous debris caniclastic sands suggest contemporaneous erosion from falling on a surface beneath this depth will be more or less higher areas. dissolved, and resulting sediment will be both thinner and There are two notable occurrences to Turonian to San- less calcareous compared to that accumulated during the tonian carbon-rich sediment (DSDP Sites 146 and 153). same interval at shallower depth. Thus, the thickness and Carbonaceous sediments of this age are otherwise unknown calcium carbonate content of sediment above the crust are in the world ocean, with the exception of a site western critical to an interpretation of the original bathymetry of the

58 THOMAS W. DONNELLY basalt mass. fied These currents can potentially tell us of the overall The present depths of the B" are an approximation of environment of deposition for the Caribbean basalt prov- the relative palaeodepths at the time of cessation of the ince. The thinness of sediment at Site 151 (Beata Ridge) can igneous event, that is, at the beginning of the post-B" sedi- be explained by its probable original shallow water depth at mentation. It is not possible to correct these depths to the the approximate eruptive centre of the plateau. Bottom true depth, because the post-igneous event thermal subsi- currents increase their velocity if the flow is constricted dence history si unknown and likely to be considerably above a topographic rise, and the thinned section at 151 is different from 'normal' oceanic crust. Further, DSDP Site probably a good example. Saunders et al.58 discussed the 151 lies close to a fault scarp and would have to be adjusted thinning of the remaining southern Venezuelan Basin sec- to take the faulting into account. Three sites have nearly tions (150 and 153) and attributed it largely to syn- or identical depths to B": 146,4700m; 153,4690m; and 150, post-depositional currents which were more erosive closer 4710m. DSDP Site 152 is shallower, 4380m. DSDP Site to the South American continental mass. This observation 151, on the Beata Ridge, has a depth of 2410m to B". If the is pertinent to the location of the basalt plateau at the time relative order of these present depths are not very different of its formation. One theory for the original location52 is that from the palaeodepths, then 146,150, and 153 are deep, and the basalt plateau was formed in the Pacific Ocean. If it was might reflect original positions on the flanks of a swollen located in the open ocean, then it would not ordinarily be basalt plateau. DSDP Site 152 would be positioned closer affected by water currents on the flanks of one side. Alter- to the summit. DSDP Site 151 is positioned even closer to nately, if the basalt plateau was close to South America at the summit. As shown on Figure 3.la, the locations of the the time of cessation of magmatism (late Cretaceous), then drilling sites with respect to the thickness of solid crust it might be expected that the effect of erosion water currents largely support these conclusions, with 146, 150, and 153 close to this land mass could be detected (as Saunders et located over thinner crust, 152 intermediate, and 151 located al.58 suggested), and the hiatuses in the B"-A" section can be where the crust is much thicker. explained by these currents. If the plateau were located far The variable carbonate contents of the supra-B" sedi- from any continental land mass in the Pacific, then this ment in these sections reveal a similar variability in the effect would not be plausible. inferred palaeodepths. DSDP Sites 146 and 153 have the Differing lithologies of the B"-A" interval are also lowest carbonate content in the Cretaceous portion of the reflected in differing acoustic velocities. Thus, Ludwig et section (Fig. 3.7). DSDP Site 152 has higher carbonate, al.46 found that the velocities of the B"-A" interval were suggesting a shallower palaeodepth. DSDP Site 150 (not tightly grouped at 2.70 ± 0.13 km s-1 for the central Vene- shown) has a highly thinned section with low carbonate zuelan Basin. Diebold et al.10 found consistently higher content. DSDP Site 151 (Beata Ridge; not shown) is highly velocities in the southern Venezuelan Basin: arranging their thinned, but has a high carbonate content. Drilling at this site five stations north to south these are 2.5,3.7,3.4,3.8, and found two hard grounds (basal Paleocene, Santonian); be- 3.5 km s-1. Stoffa et al.61 found velocities of 3.0 to 3.2 km cause of poor recovery of much of the section there may s-1 for the above mentioned line approaching DSDP Site have been more that were missed. These are thinned, cherty 153. Sediments with a lower content of carbonate from sediments that seem to have originated near the crest of a DSDP Site 146 were found to have a lower acoustic velocity. topographic elevation, where normally slow ocean currents Thus, velocity might be a proxy measurement for carbonate speed up passing over the crest and cause increased erosion. content and, indirectly, for palaeodepth. The thinning of the B"-A" section at DSDP Site 151 was In summary, the present depth, thickness, carbonate probably due to that section being located near the summit content (where known) and velocity of the B"-A" interval of the plateau, and therefore it suffered extensive post-erup- all support the following synthesis for the sedimentary tion sediment thinning by currents passing over the summit. environment of the Caribbean following the eruption of the Of further interest is the regular decline with time of the plateau basalt. Sediments deposited to the west were depos- carbonate content at DSDP Site 146 during the Cenozoic. ited near the eruptive centre (possibly near the Beata Ridge) This decrease appears to indicate a steady increase in water and highest topography, and are carbonate-rich. Flank sedi- depth on the flank, probably resulting from heat loss from ments, to the east and southeast of the Beata ridge, are the cooling plateau basalt. carbonate poor, as found by drilling and by lower acoustic As mentioned above, the thickness of the B"-A" inter - velocity. At DSDP Site 146 subsidence, pos sibly as the vals at the five DSDP sites were very different. Very likely result of heat loss, is recorded as a regularly declining these differences of thickness resulted in part from bottom carbonate content in the late Cretaceous, reaching zero in currents, which either prevented the deposition of sediment the early Cenozoic, roughly 20 Myr after cessation of mag- or removed sediment previously deposited, but not yet lithi- matism (Fig. 3.7a). At the shallowest site (151), which lies

59 The Caribbean Sea Floor

over the thickest crust, the sediment was further severely shoaling. These studies use pelagic sediment sections (ob- thinned by bottom currents, resulting in hiatuses and local tained by deep-sea drilling) as records of the palaeoenviron- hardground formation. ments of the oceans on either side of the isthmus. As the isthmus shoaled during the later Cenozoic, waters of succes- Post-A" sediments sively shallower depths were isolated on either side. For The sedimentary section of Eocene age is notably sili- example, Caribbean sediments record a decreasing supply of ceous, as are coeval sections throughout tropical oceans biological silica from the Eocene to the middle Miocene, elsewhere. The explanation for the silica content is the whereas eastern Pacific sediments record a continuing high prodious productivity of siliceous organisms at that time, supply of biological silica14. This observation is explained as which must, in turn, be related to high silica input to the a decreasing supply of deep and intermediate high-silica oceans. During the middle-late Eocene the world experi- waters from the Pacific to the Caribbean and, thus, the enced the maximum rate of siliceous deposition (and, thus, gradual rising of the Central American isthmus. The cessa- supply of silica to the oceans by rivers) in the Cenozoic and, tion of siliceous productivity at about 15 Myr represents a possibly, in the Phanerozoic. The recrystallization of these relatively deep (perhaps about 1000 m) palaeodepth for the siliceous oozes produced the excellent acoustic reflector isthmus. Keigwin38 examined the oxygen isotopes of which came to be known as A in the Atlantic, A' in the foraminiferal oozes from eastern Pacific and Caribbean Pacific, and A" in the Caribbean. A high silica content has sites. The isotopic ratios reflect the character of surface been noted in correlative on-land strata around the Carib- oceanic water and their divergence upwards, beginning at bean, notably on Puerto Rico48. However, their interpreta- 4.2 Myr, shows that the Caribbean surface waters became tion of this silica as volcanogenic is clearly not likely in the more saline than the Pacific at this time. This implies a face of such abundant biological silica in the oceanic basins. palaeodepth shallow enough to prevent extensive surface The thickness and inferred lithology of the sediments water exchange across the isthmus. Keller et al.39 studied lying above horizon A", as well as the physical character of the microfossils of sediments at similar sites. Their studies the horizon itself, provides information on the palaeodepth show faunal and floral contrasts which they similarly inter- of the Caribbean. Horizon A" is identified reliably only in pret as shoaling events at 6.2,2.4 and 1.8 Myr, each of which the Venezuelan Basin, and workers have been reluctant to implied successively shallower palaeodepths. The conclu- carry the correlation into the Colombian Basin. As noted sions of these studies reinforce the view that the Central above, A" was identified by drilling at DSDP Site 152 on American isthmus gradually shoaled and closed vertically, the tower Nicaraguan Rise. The further extrapolation of this and did not close horizontally (as a door), such as was horizon further west and south in the Colombian Basin is postulated by Wadge and Burke63. not confirmed. In the Yucatan Basin Rosencrantz did not At the present day the Caribbean is effectively isolated find a reflector that plausibly even resembles A". In the from the Atlantic, with the deepest sill about 1600 m. The Venezuelan Basin, mainly in the southern half, and in the Caribbean has distinctive deep water, which differs in char- southern Colombian Basin there are numerous records of a acter from North Atlantic deep water. Pelagic sediments horizon that resembles A" and might be considered correla- carry a built-in record of one of the most important attributes tive. The disappearance of A" to the west might result from of the ocean; the depth of the CCD, below which calcite shallower palaeodepths, at which abundant calcareous re- dissolves and calcareous sediments do not accumulate. This mains were mixed with siliceous debris. Under these cir- depth depends on the chemistry of the water column and cumstances, the later history of the radiolarian opal might presently is at about 5000 m in the north Atlantic. Through- have been to form microscopic chert nodules in a calcareous out the Atlantic, pelagic sediments record a relatively sudden section, which would not have the high acoustic reflectivity deepening of the CCD at about the Miocene-Pliocene of a pure chert layer. boundary. Caribbean pelagic sediments record this same Drilling recovery of the interval close to A" was poor. event14; at the time of the Miocene-Pliocene boundary Much of this section consists of alternating very hard and (about 5 Myr), the Caribbean was evidently broadly con- relatively soft beds, and successful recovery of (billed sedi- nected to the Atlantic at depths approaching 5000 m. The ment is very difficult in these cases. The post-A" interval is present relatively shallow sill depth between the Caribbean virtually entirely unlithified. It has yielded a clear sedimen- and the Atlantic must have formed in the last 5 Myr, possibly tary history at the two sites (29 and 149) where it was cored by compression in a north-south direction across the Carib- completely, and correlative bits and pieces of sedimentary bean. history at the remaining sites where it was spot-cored. Several studies of pelagic sediment on opposite sides of the Central American isthmus document its gradual

60 THOMAS W. DONNELLY

SUMMARY equatorial current of the tropical Atlantic to its present The bulk of our knowledge of the Caribbean crust comes northerly direction, giving us our present Gulf Stream. from geophysical studies, but has been confirmed in part by deep-sea drilling in 1970 and 1971. This body of knowledge REFERENCES is about 40 years old, making it one of the latest regions of the Caribbean for which important geological knowledge 1Bader, R.G. et al 1970. Initial Reports of the Deep Sea has been obtained Drilling Project, VolumeIV. U.S. Government Printing The Caribbean oceanic crust is one of the largest Office, Washington, D.C., 753 pp. Phanerozoic igneous provinces (plateau basalts) in the 2Biju Duval, B., Mascle, A., Montadert, L. & Wanneson, J. world. Its site is more or less astride the Jurassic mid-ocean 1978. Seismic investigations in the Colombia, Vene- ridge which separated the north Atlantic continental land zuela, and Grenada basins, and on the Barbados Ridge masses. Why it formed where and when it did (its absolute for future IPOD drilling. Geologie en Mijnbouw, 57, location remains uncertain) is unknown. There are several 106-116. other, similar occurrences in the world (several are in the 3Bowin, C.O. 1968. Geophysical study of the Cayman western Pacific) of this approximate age (Aptian or Albian Trough. Journal of Geophysical Research, 73, 5159- to Turoman). At the moment our only firm conclusion is that 5173. this period of earth history recorded immense mantle melt- 4Bowland, C.L. & Rosencrantz, E. 1988. Upper crustal ing over a wide area. It might be sugges ted that the ancestral structure of the western Colombian Basin, Caribbean Caribbean was located adjacent to one of the western Pacific Sea. Geological Society of America Bulletin, 100, 534- plateau basalts of about this age, such as the Ontong-Java 546. Plateau, but there is no good evidence for this idea. 5Burke, K.P., Fox, P.J. & Sengor, A.M.C. 1978. Buoyant The Caribbean oceanic crust unifies the geologic his- ocean floor and the evolution of the Caribbean. Journal tory of the entire province. Without consideration of the of Geophysical Research, 83, 3949-3954. crust, one can hardly relate the geologies of, say, Hispaniola 6Case, J.E. & Holcombe, T.L. 1980. Geologic -tectonic map and Curasao, or Puerto Rico and Costa Rica. When one of the Caribbean region, scale 1:2,500,000. U.S. Geo- places the Caribbean crust in a broader context, the unity of logical Survey Miscellaneous Investigations, Map I- the entire region becomes evident. Most of the Caribbean 1100. island, and possibly some of the continental areas (espe- 7Case, J.E., MacDonald, W.D. & Fox, PJ. 1990. Caribbean cially northern South America) formed peripherally to the crustal provinces; seismic and gravity evidence: in vast basalt plateau that dominated this area in the late Dengo, G. & Case, I.E. (eds), The Geology of North Cretaceous. Whether this plateau formed well out in the America, Volume H, The Caribbean Region, 15-36. Pacific, close to the 'jaws' of Central America, or was Geological Society of America, Boulder. already close to its present, enclosed position is not at all 8CAYTROUGH, 1977. Geological and geophysical inves- clear, but there are two types of evidence from the supra- tigations of the Mid Cayman Rise spreading center: crustal sediment cover that the plateau was not located far initial results and observations: in Talwani, M., Harri- from the continent. The first is the occurrence of high-carbon son, C.G. & Hayes, D.E. (eds), Deep Drilling Results sediments, which imply local restriction ofbottom waters in the Atlantic Ocean: Ocean Crust, 66-93. American producing anoxia. The second is the record of hiatuses at Geophysical Union, Washington, D.C. sites 150 and 153, implying that deep currents flowing close 9Dengo, G. & Case, J.E. (eds). 1990. The Geology of North to the South American continental slope attained a higher America, Volume H, The Caribbean Region. Geological velocity due to topographic constrictions, and in turn eroded Society of America, Boulder, 528 pp. much of the sediment cover or prevented its accumulation 10Diebold, J.B., Stofla, P.L, Buhl, P. & Truchan, M. 1981. in the first place. Venezuelan Basin crustal structure. Journal of Geo- Post-A" sediment records the closure through shoaling physical Research, 86, 7901-7923. of the Central America isthmus and, with it, the closure of 11Donnelly, T.W. 1973a. Late Cretaceous basalts from the the broad oceanic connection between the Atlantic and Caribbean, apossible flood basalt province of vast size. Pacific Oceans. The consequences of this closure have been EOS, 54,1004. felt elsewhere, as in the invasion of land mammal faunas 12Donnelly, T.W. 1973b. Magnetic anomaly observations in from South to North America and, more spectacularly, vice the eastern Caribbean Sea: in Edgar, N.T., Saunders, J.B. versa. Still largely unexplored are the potentially immense palaeoclimatic implications of a major diversion of the north et al (eds), Initial Reports of the Deep Sea Drilling Project, Volume XV, 1023-1030. U.S. Government

61 The Caribbean Sea Floor

Printing Office, Washington, D.C. Gulf of Mexico. Journal of Geophysical Research, 65, 13Donnelly, T.W. 1989. Geologic history of the Caribbean 4087-4126. and Central America: in Bally, A.W. ft Palmer, A.R. 24Ewing, J.I., Edgar, N.T. & Antoine, J. 1970. Structure of (eds), The Geology of North America, Volume A, The the Gulf of Mexico and Caribbean Sea: in Hill, M.N. Geology of North America—An Overview, 299-321. (ed.), The Sea, Volume 4,321-358. Wiley, New York. Geological Society of America, Boulder. 25Ewing, J.I., Talwani, M., Ewing, M. & Edgar, N.T. 1967. 14Donnelly, T.W. 1990a. Pelagic sediment, deep water Sediments of the Caribbean. Studies in Tropical Ocean- chemistry, and tectonics: an application of the history ography, 5, 88-102. of biological sediment accumulation on the tectonic 26Ewing, J.I. & Tirey, G.B. 1961. Seismic profiler. Journal history of the Caribbean. Rivistaltaliana di Paleontolo- of Geophysical Research, 66, 2917-2927. gia e Stratigrafia, 96,143-164. 27Fox, P.J. & Heezen, B.C. 1985. Geology of the Caribbean 15Donnelly, T.W. 1990b. Caribbean biogeography: geo- crust: in Nairn, A.E.M. & Stehli, F.G. (eds), The Ocean logical considerations on the problem of vicariance vs Basins and Margins, Volume 3, The Gulf of Mexico and dispersal: in Azarolli, A. (ed.), Biogeographical As- Caribbean, 421-466. Plenum, New York. pects of Insularity. Accademia Nazionale dei Lincei, 28Ghosh,N., Hall, S.A. & Casey, J.F. 1984. Seafloor spreading Atti dei Convegni Lincei, 85, 595-609. anomalies in the Venezuelan Basin. Geological Society 16Donnelly, T.W., Beets, D., Carr, M.J., Jackson, T., Klaver, of America Memoir, 162, 65-80. G., Lewis, J., Maury, R., Schellekens, H., Smith, A.L., 29Hall, S.A. & Yeung, T. 1980. A study of magnetic anomalies Wadge, G. ft Westercamp, D. 1990. History and tec - in the Yucatan Basin: in Snow, W., Gil, N., Llinas, R., tonic setting of Caribbean magmatism: in Dengo, G. & Rodriguez-Torres, R., Seaward, M. & Tavares, I. (eds), Case, J.E. (eds), The Geology of North America, Vol- Transactions of the Ninth Caribbean Geological ume H, The Caribbean Region, 339-374. Geological Conference, Santo Domingo, Dominican Republic, Au- Society of America, Boulder. gust 16th-20th, 1980, 2, 519-526. 17 Donnelly, T.W., Melson, W., Kay, R. & Rogers, J.J.W. 30Hayes, D.E., Pimm, A.C. etal. (eds). 1972. Initial Reports of 1973. Basalts and dolerites of late Cretaceous age from the Deep Sea Drilling Project, Volume XIV. U.S. the Central Caribbean: in Edgar, N.T., Saunders, J.B. et Government Printing Office, Washington, D.C., 975 pp. al. (eds), Initial Reports of the Deep Sea Drilling Pro- 31Hersey, J.B. 1963. Continuous seismic profiling: in Hill, ject, VolumeXV, 989-1012. U.S. Government Printing M.N. (ed.), The Sea, Volume 3, 47-72. Wiley, New Office, Washington, D.C. York. 18 Edgar, N.T., Ewing, J.I. ft Hennion, J. 1971. Seismic 32Holcombe, T.L., Ladd, J.W., Westbrook, G., Edgar, N.T. & refraction and reflection in the Caribbean Sea Ameri- Bowland, C.L. 1990. Caribbean marine geology; ridges can Association of Petroleum Geologists Bulletin, 55, and basins of the the plate interior: in Dengo, G. & Case, 833-870. 19 J.E. (eds), The Geology of North America, Volume H, Edgar, N.T., Holcombe, T.L., Ewing, J.I. ft Johnson, W. The Caribbean Region, 231-260. Geological Society of 1973. Sedimentary hiatuses in the Venezuelan Basin: America, Boulder. in Edgar, N.T., Saunders, J.B. et al. (eds), Initial Re- 33Holcombe, T.L. & Sharman, G.F. 1983. Post-Miocene ports of the Deep Sea Drilling Project, Volume XV, Cayman Trough evolution: a speculative model. Geol- 1051-1062. U.S. Government Printing Office, Wash- ogy, 11, 714-717. ington, D.C. 34 20 Holcombe, T.L., Vogt, P.R., Matthews, J.E. & Murchin- Edgar, N.T., Saunders, J.B. et al. (eds). 1973. Initial son, R.R. 1973. Evidence for sea-floor spreading in the Reports of the Deep Sea Drilling Project, Volume XV. Cayman Trough. Earth and Planetary Science Letters, U.S. Government Printing Office, Washington, D.C., 20,357-371. 1137 pp. 35 21 Hopkins, H.R. 1973. Geology of the Aruba Gap abyssal Epp, D., Grim, P J. ft Langseth, M.G. 1970. Heat flow in plain near DSDP site 153: in Edgar, N.T., Saunders, the Caribbean and Gulf of Mexico. Journal of Geo- J.B. et al. (eds), Initial Reports of the Deep Sea Drilling physical Research, 75, 5655-5669. 22 Project, Volume XV, 1039-1050. U.S. Government Erickson, A.J., Helsey, C.E. & Simmons, G. 1972. Heat Printing Office, Washington, D.C. flow and continuous seismic profiles in the Cayman 36Houtz, R.E. & Ludwig, W.J. 1977. Structure of the Co- Trough and Yucatan Basin. Geological Society of lombian Basin, Caribbean Sea. Journal of Geophysical America Bulletin, 83, 1241-1260. 23 Research, 82, 4861-4867. Ewing, J.I., Antoine, J. & Ewing, M. 1960. Geophysical 37Kafka, A.L. & Weidner, D.J. 1979. The focal mechanisms measurements in the western Caribbean Sea and in the

62 THOMAS W. DONNELLY

and depths of small earthquakes as determined from Damond, P. 1979. Upraised Caribbean sea floor below Rayleigh-wave patterns. Seismological Society of acoustic reflector B" at the southern margin of Haiti. America Bulletin, 69, 1379-1390. Geologie en Mijnbouw, 58, 71-83. 38Keigwin, L.D., Jr. 1979. Pliocene closing of the Isthmus 50Officer, C.B., Ewing, J.I., Edwards, R.S. & Johnson, H.R. of PanaMyr based on biostratigraphic evidence from 1957. Geophysical investigations in the eastern Carib- nearby Pacific Ocean and Caribbean Sea cores. Geol- bean; Venezuelan Basin, Antilles island arc, and Puerto ogy, 6,630-634. Rico trench. Geological Society of America Bulletin, 39Keller, G., Zenker, C.E. & Stone, S.M. 1989. Late Neo- 68, 359-378. gene history of the Pacific-Caribbean gateway. Journal 51Officer, C.B., Ewing, J.I., Hennion, J.F., Haikrider, D.G. of South American Earth Sciences 2, 73-108. & Miller, D.E. 1959. Geophysical investigations in the 40Kolla, V., Buffler, R.T. & Ladd, J.W. 1984. Seismic eastern Caribbean: summary of 1955 and 1956 cnisies: stratigraphy and sedimentation of the Magdelana in Ahrens, L.H., Press, F., Rankama, K. & Runcorn, Fan, southern Colombian Basin, Caribbean Sea. S.K. (eds), Physics and Chemistry of the Earth, Volume American Association of Petroleum Geologists 3, 17-109. Pergamon, New York. Bulletin, 68, 316-332. 52Pindell, J.L. & Barrett, S.F. 1990. Geologic evolution of 41Ladd, J.W., Holcombe, T.L., Westbrook, O.K. & Edgar, the Caribbean region; a plate tectonic perspective: in N.T. 1990. Caribbean marine geology: active margins Dengo, G. & Case, J.E. (eds), The Geology of North of the plate boundary: in Dengo, G. & Case, J.E. (eds), America, Volume H, The Caribbean Region, 405-432. The Geology of North America, Volume H, The Carib- Geological Society of America, Boulder. bean Region, 261-290. Geological Society of America, 53Prell, W.L.,Gardner, J.V. et al. (eds). 1982. Initial Boulder. Reports of the Deep Sea Drilling Project, Volume 42Ladd, J.W. & Watkins, J.S. 1978. Active margin struc- LXVIII. U.S. Government Printing Office, tures within the north slope of the Muertos Trench. Washington, D.C., 495 pp. Geologie en Mijnbouw, 57,255-260. 54Raff, A.D. 1973. Site 145: in Edgar, N.T., Saunders, J.B. 43Ladd, J.W. & Watkins, J.S. 1980. Seismic stratigraphy of et al. (eds), Initial Reports of the Deep Sea Drilling the western Venezuelan Basin. Marine Geology, 35, Project, Volume XV, 1063-1066. U.S. Government 21-41. Printing Office, Washington, D.C. 44Ladd, J.W., Worzel, J.L. & Watkins, J.S. 1977. 55Rosencrantz, E. 1990. Structure and tectonics of the Multifold seismic reflection records from the Yucatan Basin, Caribbean Sea, as determined from northern Venezuelan Basin and the north slope of the seismic reflection studies. Tectonics, 9, 1037- Muertos Trench: in Talwani, M. & Pitman, W.C., III 1059. (eds), Island Arcs, Deep-Sea Trenches and Back- 56Rosencrantz, E., Ross, M.I. & Sclater, J.G. 1988. Age Arc Basins, 41-56. American Geophysical Union, and spreading of the Cayman Trough as Washington, D.C. determined from depth, heat flow, and magnetic 45Lu, R.S. & McMillan, K.J. 1982. Multichannel seismic anomalies. Journal of Geophysical Research, 93, survey of the Colombian Basin and adjacent 2141-2157. 57 margins: in Watkins, J.S. & Drake, C.L. (eds), Rosencrantz, E. & Sclater, J.G. 1986. Depth and age in Studies in Continental Margins. American the Cayman Trough. Earth and Planetary Science Association of Petroleum Geologists Memoir, 34, Letters, 79,133-144. 58 395-410. Saunders, J.B., Edgar, N.T., Donnelly, T.W. & Hay, 46Ludwig, W.J., Houtz, R.E. & Ewing, J.I. 1975. Profiler- W.W. 1973. Cruise synthesis: in Edgar, N.T., Sonobuoy measurements in Colombian and Saunders, J.B. et al. (eds), Initial Reports of the Deep Venezuelan Basins, Caribbean Sea. American Sea Drilling Project, Volume XV, 1077- Association of Petroleum Geologists Bulletin, 59, 1111. U.S. Government Printing Office, Washington, 115-121. D.C. 47Matthews, J.E. & Holcombe, T.L. 1976. Regional geo- 59Schmidt-Effing, R. 1979. Alter und Genese des logical/geophysical study of the Caribbean Sea Nikoya-Komplexes, einer ozeanischen Palaokruste, (Navy Ocean areaNA-9); 1, geophysical maps of the Oberjura bis Eozan, im Siidlichen Zentralamerika. eastern Caribbean, scale 1:2,000,000. U.S. Naval Geologische Rundschau, 68,457- 494. Oceanographic Office, Reference Publication RP3. 60Silver, E.A., Case, J.E. & MacGillavry, H.J. 1975. 48Mattson, P., Pessagno, E.A., Jr. & Helsey, C.E. 1972. Geophysical study of the Venezuelan borderland. Outcropping layer A and A" correlatives in the Geological Society of America Bulletin, 86, 213-226. Greater Antilles. Geological Society of America 61Stoffa, P.L., Mauffret, A., Truchan, M. & Buhl, P. 1981. Memoir, 132, 57-66. Sub-B" layering in the southern Caribbean: the Aruba 49Maurrasse, F.J.-M., Husler, J., Georges, G., Schmitt, R. & gap and Venezuelan Basin. Earth and Planetary Sci-

63 The Caribbean Sea Floor

ence Letters, 53, 131-146. NOTE ADDED IN PROOF 62 Talwani, M, Windisch, C.C., Stoffa, P.L., Buhl, P. & 66 Subsequent to the preparation of this chapter, Bowland Houtz, R.E. 1977. Multi-channel seismic study in the published a more complete account of the seismic stratigra- Venezuelan Basin and the Curasao Ridge: in Talwani, phy and inferred geological history of the western Colom- M. & Pitman, W.C., III (eds), Island Arcs, Deep-Sea bian Basin. In this account there is discussion of the seismic Trenches and Back-Arc Basins, 83-98. American Geo- velocity of some sediment layers. physical Union, Washington, D.C. 63Wadge, G. & Burke, K. 1983. Neogene Caribbean plate rotation and associated Central American tectonic evo- REFERENCE lution. Tectonics, 2, 633-643. 64 Weidmann, J. 1978. Ammonites from the Curacao lava Bowland, C.L. 1993. Depositional history of the western formation, Curasao, Caribbean. Geologie en Colombian Basin, Caribbean Sea, revealed by seismic Mijnbouw, 57,361-364. 65 stratigraphy. Geological Society of America Bulletin, Woodring, W.P. 1954. Caribbean land and sea through the 105, 1321-1345. ages. Geological Society of America Bulletin, 65, 710- 732.

CHAPTER 4

Cuba

GRENVILLE DRAPER and J. ANTONIO BARROS

Department of Geology, Florida International University, Miami, Florida 33199, U.SA.

INTRODUCTION provinces in which geological features occur, and these are shown in Figure 4.1. Figure 4.1a shows the provinces prior THE CUBAN archipelago consists of some 4,194 islands to 1959 (as referred to in the older literature), and Figure and cays, including the main island of Cuba and the Isla de 4.1b shows how they were re-arranged following the revo- la Juventud (previously the Isla de los Pinos—the Isle of lution. Pines), covering a total of 110,922 square km (the main The physiographic relief of most of Cuba (Fig. 4.2) is island making up some 105,007 squar e km). The archipel- relatively subdued compared to other islands in the Greater ago is the only part of the Greater Antilles situated on the Antilles, and mountainous areas are found only in the west- North American Plate. ern province of Pinar del Rio, the Escambray region in the Modern geological investigations in Cuba were initi- centre of the island and in the eastern province of Oriente. ated in the 1950s by oil company geologists11,25. Much of The last contains Pico Turquino in the Sierra Maestra which, this work remains unpublished, but several review papers at 1,972 m, is the archipelago's highest peak. were published in the 1960s and 1970s16,18,33, including summaries of the accomplishments of petroleum geolo- 9,19,26 gists . Since 1959, much new knowledge has been BASIC COMPONENTS OF CUBAN GEOLOGY added by Cuban geologists aided by eastern European colleagues. There is now a large literature concerning the The geology of Cuba differs significantly from that of other geology of the island, including geologic, tectonic and met- areas of the Greater Antilles in several respects. Cuba con- allogenic maps at a scale of 1:500,000, and a new geologic tains Precambrian rocks (900 Ma metamorphic rocks in map at a scale of 1:250,00024 With the exception of the 26 17 Santa Clara province; Fig. 4.1b) and extensive outcrops of papers by Pardo and Lewis , few of the general descrip- continental margin, sedimentary rocks of Jurassic to Creta- tions of the island's geology are available in English. The ceous age. It is structurally characterized by large thrust and purpose of this chapter, therefore, is to give a broad over - nappe structures, which are not present in the other islands view of modern interpretations of Cuba's geology for an of the northern Caribbean. English speaking audience. The main text is supplemented Cuba can be divided into two broad geological prov- by two appendices, explaining the map symbols used in the inces (Fig. 4.3): Cuban geological literature (Appendix 1) and the sub-Upper Eocene stratigraphy of the structural facies zones of the four 1. Western and central Cuba constitute a complexly principal structural blocks (Appendix 2). deformed orogen resulting from the collision of amid to late Cretaceous island arc with the late Jurassic to late Creta- ceous sedimentary rocks of the Florida-Bahamas plat- GEOGRAPHY form10,34. This event occurred in the late Cretaceous, producing both the obduction of the Cuban ophiolite belt The major geographic features of Cuba are shown in Figures and a large, northward-verging, fold and thrust belt. Some 4.1 and 4.2. The Cuban literature often refers to the political of these structures were reactivated by a second orogenic

65 Cuba

Figure 4.1. Major geographic features and provinces of Cuba (a) prior to the 1959 revolution and (b) at the present day.

Figure 4.2. Physiography of Cuba. Contours above sea level are shown as solid lines at 1,000 m intervals. Stippled areas indicate low lying, wetland regions. Bathymetric contours are shown as broken lines, also at 1,000 m intervals. Adapted from Weyl35.

66 G. DRAPER and J. A.. BARROS

Figure 4.3. Main structural blocks and structural-facies zones of Cuba; Pinar del Rio Block (Fig. 4.4), Isla de la Juventud Block (Fig. 4.5), Central Cuba Block (Fig. 4.6), Camaguey Block (Fig. 4.9), and Oriente Block (Fig. 4.10). Note especially the Jurassic ophiolites, which lie at the base of the Cretaceous island arc rocks, both of which are thrust over the Jurassic to Cretaceous, passive margin, sedimentary rocks that form the northernmost structural facies zones of Cuba.

28 phase in the Paleocene to early Eocene . This later defor- 1. Pinar del Rio block (located west of Havana). mation seems to have been diachronous and becomes pro- 2. Isla de la Juventud block. gressively younger to the east . The precise plate tectonic 3. Central Cuba block (located between Havana and the La configurations that gave rise to this orogen are still hotly Trocha fault zones). debated. These orogenically deformed rocks are overlain by 4. Camaguey block (between the La Trocha and Cauto fault relatively undeformed, post-orogenic sedimentary rocks. zones). 2. Eastern Cuba (southeast of the Cauto basin) is char- 5. Oriente block (south and east of the Cauto fault zone). acterized by a Cenozoic (Paleocene-Middle Eocene) volcanic-plutonic arc complex of the Sierra Maestra. North and It is important to note that this division is slightly east of the Sierra Maestra, ophiolitic and arc rocks of the different from that used by previous authors, especially in Mesozoic orogen occur, overlain by Paleogene sedimentary our distinction between the central Cuban and Camaguey rocks and tuffs. Although the older rocks have similarities blocks. However, it is considered that there are sufficient to those in central Cuba, they are different in the sense that structural, if not stratigraphic, differences between these two continental margin rocks are rarer. In contrast to western and regions to justify this distinction. central Cuba, Tertiary sedimentation related to tectonism Within these structural blocks, the geology is further persisted until the Oligocene. sub-divided into what Cuban investigators have called structural-facies zones. The structural-facies zones can be STRUCTURAL GEOLOGY OF OROGENICALLY identified, partially or completely, in each of the blocks and DEFORMED ROCKS OF CUBA are fault bounded belts or nappes, which have distinctive stratigraphic, metamorphic and/or palaeogeographic char- The orogenically deformed rocks of Cuba (that is, pre-Mid- acteristics. The original concept was introduced by Pardo25, dle Eocene in the west and central part of the island; pre- after which it was adopted and refined by numerous Oligocene in the east) can be conveniently divided into five 8,11,26 17,19 authors . Hatten et al. have gone so far as to label major geologic -structural segments or blocks . These are, these zones in central Cuba as tectonostratigraphic terranes, from east to west (Fig. 4.3): a concept that is still controversial. The structural-facies

67 Cuba

Figure 4.4. Structural facies zones of the Pinar del Rio Block of western Cuba Stippled area represents the Zaza zone (island arc rocks and basal ophiolites) which extend west of Havana

zone concept is nonetheless usefiil in describing the major the Cayo Coco and Los Remedios zones of central Cuba (see features of the geology of Cuba and is adopted in the below). These rocks are deformed by thrust faulting into at following account. A thorough terrane analysis of Cuba least three nappes. Subsurface information suggests that awaits further investigation. they structurally overlie Upper Cretaceous to Paleocene flysch-like sandstones and shales.

PINAR DEL RIO BLOCK Sierra del Rosario zone The rocks of the Sierra del Rosario form an antiformal Western Cuba consists of five structural facies zones (Fig. arrangement of three nappes, each of which is composed of 4.4). The northernmost, which is poorly exposed, is the Jurassic to Upper Cretaceous ophiolitic rocks and (mainly) Esperanza zone. To the south, three of the zones (the Sierra carbonate sedimentary rocks. The Bahia Honda sub-zone is de Rosario in the north, the Sierra de los Organos in the south the northernmost, and structurally highest, unit of the Sierra and the Cangre zone, also in the south) form the Cordillera del Rosario zone and consists of ophiolitic rocks. The struc- de Guaniguanico (Fig. 4.2). The Cangre zone forms a thin turally lowermost units of the sub-zone consist of mafic and sliver on the southernmost flanks of the Cordillera de intermediate lavas, siliceous slates and well-laminated lime- Guaniguanico and is separated from the San Diego de los stones, whereas the uppermost units consist of basalts, gab- Banos zone by the Pinar del Rio fault. bros and ultramafic rocks. These relations have led some investigators23 to conclude that the sequence is overturned. Esperanza zone Instead, we suggest that these outcrop patterns are more The Esperanza zone outcrops as a thin belt in the likely to be due to duplication by imbricate thrusting. northernmost part of western Cuba The rocks of the region The Quinones tectonic sub-zone structurally underlies are poorly exposed and most of the useful information on the Bahia Honda sub-zone and is separated from it by a north the zone comes from subsurface studies. The rocks of this dipping thrust fault. The Quinones sub-zone consists of belt consist of Upper Jurassic to Lower Cretaceous eva- Lower Cretaceous to Maastrictian limestones organized into porites, dolomites and limestones similar to those found in three thrust sheets, each of which is associated with an

68 G. DRAPER and J. A. BARROS

Figure 4.5. Structural and metamorphic facies map of the Isla de la Juventud terrane.

olistostrome unit. of Jurassic rocks in Cuba. These are the deltaic sandstones, The Cinco Pesos tectonic sub-zone forms the south- siltstones, and micaceous and carbonaceous shales of the ward dipping, southern limb of the Rosario antiform. It Middle Jurassic (Bajocian) San Cayetano Formation. This consists of thrust sheets of Jurassic to Lower Cretaceous formation has been metamorphosed and finer grained li- carbonates with Upper Cretaceous flysch-like clastic sedi- thologies have been converted to well-developed phyllites. mentary rocks, and Jurassic deltaic sandstones (San The San Cayetano Formation is overlain by a thick Cayetano Formation). The southern limit is bounded by the sequence of Oxfordian to Tithonian shallow water lime- eastern extremity of the Pinar fault. stones of the Jagua, and the lowermost member of the Guasasa (previously Vinales) Formations. However, the Sierra de los Organos zone post-Tithonian members of the Guasasa Formation are com- The Sierra de los Organos contains the largest exposure posed of pelagic limestones and cherts, and thus record a

69 Cuba

Figure 4.6. Structural facies zones of the Central Cuba Block. Stippled area represents the Zaza zone (basal Jurassic ophiolites and Cretaceous island arc rocks). Teeth on thrust fault are on the upper plate (hanging wall) of the fault.

sudden deepening of the basin27. These in turn are (?con- in the formation have glaucophane-pumpellyite assem- formably) overlain by Upper Cretaceous to Paleocene con- blages that indicate a high pressure/low temperature meta- glomerates and sandstones. morphic environment20. Structurally, the Sierra de los Organos sequence is sliced into a series of nappes. In the northwestern Sierra

Guaniguanico the San Cayetano Formation is thrust (north San Diego de los Banos (Zaza) zone verging and north dipping) over the younger Cretaceous The San Diego de los Banos (SDB) zone occurs south carbonates. In the southeastern part of the Sierra, the San of the Cangre belt and is separated from the latter by the Cayetano Formation structurally overlies the limestones and is Pinar fault. As the SDB zone consists of a thick Tertiary separated from them by a southward-dipping thrust. The basin, and lies topographically lower than the Cordillera carbonates are thus exposed as a tectonic window in the Guaniguanico, this suggests a considerable dip-slip compo- central part of the Sierra. nent of displacement on the Pinar fault, although aleft lateral strike slip component has also been suggested. Cangre zone Limited outcrop of pre-Tertiary rocks adjacent to the The Cangre belt forms a narrow, wedge-shaped belt on Pinar fault expose Cretaceous strata with tuffaceous and the southern boundary of the Guaniguanico massif. It is epiclastic layers. The presence of these rocks suggests that composed of quartzose meta-arenites, mica phyllites and the basement of the San Diego de los Banos basin is correlatable with the Cretaceous island arc rocks (Zaza zone) of occasional graphitic phyllites. According to Millan and 13 Somin , these rocks are metamorphosed equivalents of the central Cuba . San Cayetano Formation. The metamorphic grade is difficult to estimate, but sills of metadolerite and gabbro found

70 G. DRAPER and J. A. BARROS

Figure 4.7. Schematic cross section across central Cuba (after Lewis17). See Figure 4.3 for approximate location.

Some small plutons of unknown age intrude the meta- ISLA DE LA JUVENTUD (ISLA DE morphic rocks in the northern part of the island. A small area

PINOS) BLOCK of poorly exposed (and poorly dated) Cretaceous volcanic Most of the Isla de la Juventud consists of metamorphosed rocks outcrops in the northwest. Pliocene to Quaternary Mesozoic sedimentary rocks (Fig. 4.5). The lower part of deposits compose the low-lying southern third of the island. the protolith sequence (Canada Formation) incorporates intercalations of micaceous and graphitic pelites with quartzose psammites. The middle part of the sequence CENTRAL CUBA BLOCK (Agua Santa Formation) contains similar lithologies, but is characterized by increasing quantities of marble and calc - Central Cuba is divided into several structural facies zones silicate beds. The upper part (Gerona Group) consists almost (Fig. 4.6). Although most authors agree on the geology that entirely of black to dark grey dolomitic marbles. A scarce occurs in these zones, several nomenclatural schemes have and poorly preserved fauna indicates a Jurassic age and, been proposed which has led to some confusion in the therefore, the rocks may be correctable with the Jurassic literature. Here we use the scheme of Meyerhoff and Hat- San Cayetano and Cangre zone of the Cordillera de ten19 and Hatten et al.12. Figure 4.7 shows schematic cross Guaniguanico (G. Millan, personal communication, 1990). sections across central Cuba and illustrates the broad struc- Overall, the metamorphism of the Jurassic rocks of Isla tural relations between the zones. The first four of the zones de la Juventud is of high temperature/medium pressure type, described below are composed mainly of continental mar- which contrasts with that of the Escambray and Purial gin, continental slope and pelagic sedimentary deposits. regions (see below). Millan20 divided the rocks of the ter - rane into 5 metamorphic zones, which are, in order of Cayo Coco zone increasing metamorphic grade: greenschist (biotite zone) The Cayo Coco zone has very limited exposure and facies; staurolite-, kyanite- and garnet-bearing schists; most of the information concerning this zone comes from staurolite and kyanite schists with occasional andalusite; subsurface data derived from petroleum exploration wells. garnet, kyanite and biotite schists with occasional silliman- The oldest unit penetrated consists of Lower to Upper ite, and scapolite-bearing marbles; and sillimanite, garnet, Jurassic evaporites of the Punta Allegre Formation (these pottassium-feldspar gneisses andmigmatites. Potassium-ar - evaporites have developed diapiric structures at several gon ages from muscovite indicate that the time of metamor- localities in north central Cuba). The evaporite sequence is phism of the Isla de la Juventud massif was about 55-66 Ma overlain by Upper Jurassic to Cretaceous (Albian) shallow (late Cretaceous to early Tertiary32).

71 Cuba

Figure 4.8. Geological map of the Escambray, Mabujina-Manicaragua and southernmost Zaza zones (from Lewis17). water dolomites, anhydrites and oolitic limestones. These in the evaporites, with considerable evidence of thrust fault- are overlain in turn by deeper water, intercalated shales, ing. limestones and cherts of Aptian to Coniacian age. A major unconformity separates these units from Upper Maas- Remedios zone tnchtian to Middle Eocene marls, limestone breccias and This zone consists of a carbonate bank sequence of shallow water platform limestones. Structurally, the Cayo Upper Jurassic to Santonian age. This in turn is overlain by Coco zone is characterized by diapiric structures developed a thick sequence (about 2,000 m) of Maastrichtian to Paleo-

72 G. DRAPER and J. A.. BARROS

Figure 4.9. Structural facies zones of the Camaguey block. Stippled area indicates Jurassic ophiolite and Cretaceous andesitic rocks of the Zaza zone. cene limestone breccias and shallow water limestones simi- thick sequence of olistostromic conglomerates ('wild- lar to those of the Cayo Coco zone. A clastic unit consisting flysch') considered to be upper Middle Eocene. Most Cuban of Lower Eocene greywackes also occurs. High angle thrust geologists consider this deposit to mark the final movements faults, often with a right lateral strike slip component, cut of the Cuban orogeny. The deformation in the Camajuani through the limestones of the Remedios zone. zone is intense, consisting of tight, often steeply plunging folds. Both thrust and belt orthogonal tear faults are present. Camajuani (Zulueta) zone The lower part of the sedimentary sequence of the Placetus (Las Villas) zone Camajuani zone (Zulueta zone of Hatten et al ) consists of The Placetus zone (Las Villas zone of Hatten et al12) a continuous sequence of Upper Jurassic to low Upper occurs as sporadic outcrops from near Havana (the Martin Cretaceous, deep-water limestones with intercalations of Mesa tectonic window) to central Camaguey and often clastic limestones. These clastic limestones contain a shal- occupies tectonic windows beneath the allochthonous low-water fauna and were presumably washed in from a ophiolites of the Zaza zone. There are many lateral vari- nearby shallow carbonate bank (probably the Remedios ations in the deposits of this belt. In some places, the base zone). of the sedimentary sequence consists of quartzose and ark- An angular unconformity separates the lower sequence osic sandstones and conglomerates which are possibly Up- from an overlying Maastrichtian to Middle Eocene sequence per Jurassic; in others the base consists of Tithonian (Upper of polymictic olistostromic conglomerates. Clasts are de- Jurassic) radiolarian, pelagic limestones. Overlying this is a rived from the pelagic limestones of the Las Villas zone as series of Lower Cretaceous pelagic cherts, limestones and well as shallow -water clasts from the Remedios zone (the shales. Maastrichtian breccias were disconformably(?) de- latter make up about 70% of the clasts). The conglomerate posited on the Lower Cretaceous pelagics. is overlain in turn by a thinner sequence of Middle Eocene Ophiolitic rocks occur in parts of the zone, and some calcarenite and coquina beds. The final unit represented is a controversy exists concerning the age and nature of these

73 Cuba

Figure 4.10. Structural facies zones of eastern Cuba. Stippled area indicates Jurassic ophiolitic and Cretaceous andesitic rocks of the Zaza zone. Metamorphic and magmatic zones of the Cretaceous Purial and Sierras Maestra zones (respectively) are also shown.

units. Some ophiolitic rocks are clearly part of the Zaza zone The lowest structural unit recognized in the Zaza zone (see below), and are cut by dykes and granitoid stocks that outcrops its the northern margin, adjacent to the Las Villas have an island arc geochemical signature12. However, some thrust (which separates the Zaza zone from the Placetas authors have suggested that a second basaltic sequence that zone). This unit is a highly sheared serpentinite melange is not cut by granitoids may occur beneath the Jurassic containing high pressure/low temperature schists and eclo- sedimentary sequence. This has been interpreted as the gites. Lying above it is an interlayered serpentinite, gabbro basaltic portion of ocean crust. J.A.B. is doubtful of the and dolerite complex. Somin and Millan33 reported a K-Ar existence of this sec ond proposed basaltic occurrence. age of 160 Ma (middle Jurassic) from an anorthosite in this A remarkable metamorphic, pre-Mesozoic basement is complex. These two units are part of the Cuban ophiolite exposed in at least three areas of the Placetas zone, and belt which characterizes northern Cuba. We interpret this consists of marbles, mylonitic schists and granitoids. Somin unit as the oceanic basement of the overlying island arc and Millan33 reported controversial K-Ar mica ages of 910 volcanic sequence. and 945 Ma from schists near to the village of Sierra Overlying the mafic and ultramafic rocks is a pile of Morena. However, these Precambrian ages were confirmed porphyritic, basaltic and andesitic lavas (occasionally pil- by Renne et al.29, who obtained 40Ar/39Ar ages of 903 Ma lowed), overlain in turn by a sequence of tuffs and epiclastic from phlogopite in marbles of the Socorro Complex situated sedimentary rocks with interbedded pelagic and shallow- in the northwestern part of the Placetas zone. These very old water rudist limestones. The age of this volcanic -sedimen- rocks are intruded by granitoid plutons of early Cretaceous tary sequence is Aptian-Albian to Campanian4,26. This age29,33. Deformation within the Placetas zone is complex interval represents the period of island arc magmatism in the and is characterized by tight isoclinal folds and repetition of Zaza zone thoughout Cuba. Uppermost Cretaceous to Pa- sequences by steeply dipping thrust faults. leogene sedimentary rocks overlie the volcanic sequence in the Zaza zone, and are composed mainly of flyschoid grey- Zaza zone wackes and calcarenites with localized conglomerate and In contrast to the continental margin or 'miogeosyncli- breccia deposits. nal’ deposits of the structural-facies zones of north-central Cuba (see above), the Zaza zone contains Jurassic to mid- Mabujina-Manicaragua zone Cretaceous igneous rocks of oceanic and island arc ('eugeo- The Mabujina complex, a large mass of amphibolite synclinaT) origin. Although originally defined for central which is intruded by numerous granitoid bodies, lies to the Cuba, many Cuban geologist use the term Zaza zone to refer north of the Escambray complex (Fig. 4.8). In addition, to all pre-middle Cretaceous igneous rocks throughout Mabujina amphibolites also occur at the southwestern rim Cuba. of the Trinidad dome and the eastern part of the Sancti

74 G. DRAPER and J. A. BARROS

Figure 4.11. Geological map showing details of the Purial zone.

Spintus structure. This suggests that the Mabujma complex amphibolite. Mapping indicates that it is also intrusive into may have structurally overlain the entire Escambray com- Middle Cretaceous sedimentary rocks of the Zaza zone, plex prior to the doming. Whole rock K-Ar ages of 76-89 indicating a late Cretaceous intrusive age. This is also indi- Ma have been obtained, although the younger end of this cated by isotopic data. A U-Pb zircon date of 89 Ma range may represent cooling ages after the intrusion of the (Turonian) was reported by Hatten et al.12, who considered Manicaragua granitoids (see below). The age of the protolith this to represent the intrusion age of the body. K-Ar ages on of the Mabujina complex is unknown. The Mabujina amphi- biotite and hornblende are 69-73 and 69-95 Ma, respec- bolites have been interpreted as the oceanic basement of the tively. These may represent cooling ages. Strontium isotope Zaza arc (Millan and Iturralde-Vinent, personal communi- studies show that initial Sr isotope ratios are all below cations), an opinion with which we concur. If this is correct, 0.7040, indicating no continental contamination for the the Mabujina complex is perhaps correlatable with the Ju- Manicaragua granitoids (that is, an intra-oceanic island arc rassic mafic and ultramafics rocks that outcrop in the north- origin). ernmost parts of the Zaza zone. The Manicaragua unit proper is a granitoid batholith Escambray Complex that both intruded, and outcrops to the north of, the Mabujina The Escambray complex occurs as two structural

75 Cuba

Figure 4.12. Sketch section from the eastern Sierra Maestra to the Mayari-Moa-Baracoa zone (after Cobiella, personal communication). See Figure 4.3 for approximate location. Not to scale.

Figure 4.13. Oblique, sinsistral rifting of North and South America occurred in the middle Jurassic to form the Cuban passive margin on the southern boundary of North America. domes or 'cupulas' (Fig. 4.8), that is, the western Trinidad Halten et al.12 reported an 85 Ma age from an eclogite (but dome and the eastern Sancti Spiritus. Lithologically, the give no details of the determination). Metamorphism seems to complex consists of pelitic and psammitic schists and mar - have taken place sometime during the Turonian to Cam- bles, which represent a metamorphosed terrigenous carbonate panian interval. terrane. Exotic lenses and pods of serpentinized peridotite, eclogite and garnet blueschist occur within the sedimentary rocks around the margins of both domes. Am- CAMAGUEY BLOCK—EAST CENTRAL CUBA monite and radiolarian faunas discovered in less deformed parts of the Escambray complex indicate a late Jurassic The Camaguey block is bounded on the west by the La (Oxfordian) age for the sedimentary protolith of the com- Troc ha fault and the east by the Cauto Fault (Figs 4.3, 4.9). plex, that is, about the same age as the San Cayetano Although there are similarities with the geology of central Formation of western Cuba (Millan, personal communica- Cuba, there are also some significant differences. For example, tion). Granitoid intrusions do not occur in the central parts of the Escambray, Manicaragua-Mabujina, Camajuani and Hie Escambray complex. Placetas zones do not have significant outcrops in Cam- The metamorphism of the Escambray complex is in- aguey (although the Camajuani and Placetas zones are en- verted and pre-dated the doming of the complex22, and countered in windows in the ophiolite and in the therefore a concentric pattern was produced by the zoning. subsurface), and the width of outcrop of the Zaza zone is Thus, the lowest grade rocks are also structurally the lowest much greater than in central Cuba. and are found at the centre of the domes, with the highest grades being encountered at the margin. Cayo Coco and Remedios zones A mica K-Ar age of 80 Ma22 from a pegmatite cutting The Cayo Coco and Remedios zones outcrop in the the complex gives a minimum age for metamorphism, and Sierra de las Cubitas in northern Camaguey and the strati-

76 G. DRAPER and J. A. BARROS

Figure 4.14. Tectonic model of the collision of a north-facing island arc (Zaza zone) with the Cuban passive margin, (a) The Cuban passive continental margin in the late Jurassic -early Cretaceous, (b) The approach of the island arc in the middle Cretaceous, (c) Approximate configuration after the late Cretaceous -early Tertiary collision. In this model, the Zaza zone is considered totally allochthonous and overrides the passive margin; the Cuban ophiolite belt is interpreted as the forearc and/or oceanic basement of the arc. graphy of the zones is essentially the same as in and southern belts. The northern belt contains bi- central Cuba. The main difference between the two areas modal plutons of early Cretaceous age, but the southern is the age of the Cenozoic deformation. In central Cuba, belt ranges in age from Coniacian to late Campanian this deformation began in the early Eocene, but in (Stanek, personal communication). Camaguey it was initiated in the middle Eocene. Paleogene Orogen

Zaza zone The El Cobre Group is a sequence of Paleocene to The Zaza zone is separated from the continental Middle Eocene andesitic volcanic and sedimentary rocks margin sedimentary rocks to the north by a serpentinite that outcrop extensively in the Sierra Maestra of southeast- melange, as in central Cuba. The stratigraphic range of ern Cuba Extensive outcrops of the El Cobre Formation the volcanic units is essentially the same as for central also occur in the southern part of the Camaguey block. Cuba. However, the outcrop of both the ophiolitic and According to Mossakovsky et al 24, the largest area is occu- arc parts of the Zaza zone in Camaguey are broader pied by the limestones and clastic sedimentary rocks, which than in central Cuba. Another major difference are distal to the main volcanic locus and are probably also of Camaguey is the presence of several large equivalent to the upper part of the group of the Sierra plutons which intrude the volcanic rocks in northern

77 Cuba

Figure 4.15. Tectonic model of the collision of the Escambray terrane with an initially south-facing island arc, followed by emplacement of the back arc and arc onto the Cuban passive margin, (a) The approach of the Escambray terrane in the mid Cretaceous, (b) Collision and attempted subduction of the Escambray terrane with the Zaza arc in Campanian time, causing high pressue metamorphism of the former. The collision may have blocked the subduction zone and initiated a short period of underthrusting of the back arc basin beneath the arc which finally resulted in (c) the emplacement of the arc and ophiolite on the passive margin. In this model, the Cuban ophiolite belt is interpreted as part of the back arc basin and/or the oceanic basement of the rear of the arc.

Maestra. The lower, volcanic part of the sequence is found geology of the area into three structural-facies zones (Fig. only in the easternmos t part of Camaguey. 4.10), the first two of which form part of the Mesozoic orogen, and because of their dominantly igneous nature might be regarded as part of the Zaza zone2. ORIENTE BLOCK Purial zone The geology of Oriente differs from central Cuba in three The Purial zone consists of three units of regionally major respects: except for a small outcrop of metamorphic metamorphosed rocks that outcrop in the eastern part of rocks, no rocks of continental provenance are known; in Cuba and is made up of several units (Fig. 4.11). In the addition to rocks of the Mesozoic orogen, eastern Cuba also extreme east, a complex of black marbles and sericite schists contains rocks formed in a Paleogene island arc setting, occurs. On the basis of their lithology, these rocks have been which is here called the Paleogene orogen (Iturralde-Vinent, correlated with those of the Escambray region; conse- personal communication); and diastrophism and diastrophi- quently, a protolith age of Jurassic -early Cretaceous has cally controlled sedimentation may have continued to the been suggested for them, but no direct evidence has yet been Oligocene or younger. discovered for this date. A major fault separates the marble- For the purposes of this discussion we have divided the schist complex from the Purial Complex to the west. A small

78 G. DRAPER and J. A. BARROS

Figure 4.16. Palaeogeography of Cuba in the Miocene to Pliocene. complex of garnet-amphibolites, ultramafics, gabbros and lavas, tuffs and volcanogenic sedimentary rocks with occa- dolerites occurs to the west of the fault, and may compose sional limestones. Microfossils in the sedimentary rocks part of the Purial Metamorphic Complex proper. indicate a Campanian age. Rocks in the eastern and north- The Purial Metamorphic Complex consists of a large eastern part of the complex are barely metamorphosed, and area of high-pressure, low-temperature blueschist and metamorphic grade generally increases toward the south- greenschist facies rocks. Most of the protolith consists of west. Unmetamorphosed Maastnctian age rocks of the Rio 79 Cuba

Cana Formation unconformably overlie the Purial rocks, Formation) and pelagic to neritic limestones (Puerto Boni- suggesting that burial, metamorphism and subsequent un- ato Formation). On the south coast of Oriente, the lower part roofing of the complex must all have occurred in the Cam- of the El Cobre Group is intruded by low potassium grani- panian to Maastrictian, an interval of less than 20 million toids (Fig. 4.11). K-Ar ages of 46 and 58 Ma indicate an years. early to middle Eocene intrusive age. The Sierra del Convento melange, at the southwest corner of the Purial complex, is a serpentinite matrix melange and contains blocks of blueschist and amphibolite. TECTONIC EVOLUTION OF CUBA The blocks have not been dated and their age is unknown. An understanding of the tectonic evolution of Cuba in terms Mayari-Moa-Baracoa zone of the plate dynamics of the northern Caribbean is obviously This zone outcrops in the northern part of Oriente crucial to any explanation of the tectonic evolution of the province and is dominated by the two large ophiolite bodies Caribbean region as a whole. Despite progress in under- of Nipe-Cristal and Moa-Baracoa (Figs 4.10, 4.12). Neither standing the timing of events, many aspects of the tectonic of these bodies exhibits a complete ophiolite sequence as evolution of Cuba, especially the geometry and kinetics they consist mainly of serpentized harzburgite (with minor of the various crustal movements, remain elusive. Iherzolites and wherlites) with gabbros and dolerite. The age It is clear from the above discussion that Cuba com- of these bodies is not known, but they are thought to be prises five major components: Upper Jurassic to Lower Cretaceous. The Nipe-Cristal body is associated with a high pressure/low temperature ser - (1) A Jurassic to Lower Cretaceous continental margin and pentinite melange which outcrops to the south of the slope sequence deposited on both continental and oce ophiolite (Cobiella, personal communication). anic crust (western Cuba, Cayo Coco/Remedios/Zu- The oldest Cretaceous rocks in this zone are poorly lueta/Placetas zones of central Cuba and Camaguey). dated Albian?-Turonian mafic volcanic rocks of the Santo (2) A Lower to low Upper Cretaceous island arc complex Domingo Formation. These are considered by many Cuban built upon a Jurassic oceanic basement (Zaza zone geologists to be equivalents of the weakly metamorphosed throughout the island). part of the Purial Complex. (3) Jurassic continental slope deposits that outcrop south of Upper Campanian to Maastrichtian serpentinite con- the island arc complex (Escambray and Isla de la Juven- glomerates (La Picota Formation) overlie the volcanic se- tud). quence, and these are thought to have been deposited as (4) A Paleogene island arc complex (Sierra Maestra). olistostromes in front of the northward advancing Nipe- (5) Middle Eocene to Recent, post-orogenic sedimentary Cristal ophiolite thrust sheets; they therefore time the age of deposits. emplacement of the ophiolite. The conglomerates interdigi- tate with, and are ultimately overlain by, Upper Campanian Any explanation of the tectonic evolution of Cuba to Maastrichtian (possibly lowermost Paleocene), rhythmi- must explain the origin and present structural states of these cally-bedded sandstones and siltstones (Micara Formation). units. The rocks overlying the Micara Formation are pyroclas tic and sedimentary rocks are associated with the Paleogene The Mesozoic Orogen—-western and central Cuba orogen. Continental margin deposits began to accumulate in Cuba during the subsidence resulting from the rifting of Paleogene Orogen—Sierra Maestra zone South from North America during the Jurassic (the earliest The Sierra Maestra was the locus of Paleocene to mid - deposits were probably the deltaic sediments of the San dle Eocene island arc magmatism. Island arc magmatism Cayetano Formation). The age of the basement of this ceased at the end of the Campanian elsewhere in Cuba, and passive margin is uncertain. The Socorro complex29 is in this sense the geology of southern Oriente is more like Grenville in age (approximately 1,000 Ma, the same as that that of northern Hispaniola, where island arc magmatism of a large area of the North American continent east of the also persisted until the middle Eocene. present Appalachian mountain chain) and this may suggest The single most important unit in this zone is the El that at least part of the Cuban basement is Grenvillian. Cobre Group (previously Cobre Formation) which consists Alternatively, the contintental basement may have a Pan- of a thick sequence of waterlain tuff and agglomerate depos- African age (400-500 Ma) as suggested by the age of base- its with subordinate lavas and sedimentary rocks. The upper ment encountered at the base of an ODP hole drilled in the part of the El Cobre Group consists of greywackes (San Luis Gulf of Mexico northeast of Cuba. Further lithospheric extension between North and

80 G. DRAPER and J. A. BARROS

South America resulted in the formation of oceanic crust ocean floor. between the two continents (the ‘Proto-Caribbean' of Pin- As with the first model, there are problems. The model dell, Chapter 2, this volume). On the northern boundary, a implies that the Isla de Juventud and Escambray are separate passive continental margin sequence developed. Carbonate continental fragments, and while this may be the case, their banks accumulated on the subsiding, stretched continental tectonic provenance and evolution is unclear. There is little crust and deeper water, continental slope deposits probably evidence of volcanic material in the sedimentary rocks of prograded southward onto oceanic crust. the northern passive margin sequences before the Cenoma- In the early Cretaceous, an island arc system (Zaza nian27, which is at least 20 Ma after the beginning of arc zone) began to develop somewhere south or west of the activity. It is also difficult to explain the presence of high continental margin, possibly in the Pacific realm (Pindell, pressure/low temperature metamorphism in the Cangre Chapter 2, this volume). In about the Campanian this arc zone. system collided with the continental margin and began to create the Cuban orogen. The evidence for this comes from Paleogene Orogen—eastern Cuba the Campanian metamorphic ages of the Escambray (and Subduction-related magmatic activity apparently possibly the Purial) metamorphics, and the presence of ceased with this Campanian collisional event, as no evi- clasts of volcanic rock in Cenomanian passive margin sedi- dence of Maastrichtian magmatism is known in Cuba. Mag- mentary rocks. However, the exact geometry of this colli- matic activity resumed in the Paleocene, but only in eastern sion is obscure and two possible scenarios have been Cuba, where the locus of activity was the Sierra Maestra, proposed: although some pyroclastic deposits are found in the south- 1. The arc faced north (that is, the suduction zone eastern Camaguey and northern Oriente. The Paleogene dipped south), and ocean floor was subducted between the magmatic terrane terminates abruptly at the Cayman fault passive margin and the arc (Fig. 4.13,4.14). When all of the (the northernmost fault of the northern Caribbean plate ocean had been subducted, the arc collided with the passive boundary zone—NCPBZ) on the southeastern coast of margin, and completely overode it. In this scenario, the Zaza Cuba. This has led to the speculation by several authors1,3,5-7 zone is completely allochthonous, and the Escambray re that the Paleogene arc of Cuba was related to, and contigu- gion represents the structurally imbricated edge of the pas ous with, similar rocks in Hispaniola, and that latest sive margin whose inverted metamorphism resulted when Paleogene to Neogene left-lateral strike-slip on the NCPBZ the arc overode the complex. The Isla de la Juventud would has subsequently separated the two regions (that is, dis- presumably have been overridden in the same event. The persed the Paleogene terrane). ophiolites of central and western Cuba would represent An explanation of the Paleogene arc (orogen) remains fragments of the oceanic basement beneath the arc and/or an outstanding problem of plate geometry and kinematics. the Proto-Caribbean ocean crust. The Paleogene arc was deposited upon the Mesozoic oro- There are several problems with this model. It fails to genic rocks of Oriente which seem to have been already explain the location, metamoiphism (high temperature/low attached to the southern margin of the North American plate. pressure) and plutonism in the Isla de la Juventud. More- To produce arc activity means the subduction of significant over, an entire island arc would have to be detached and quantities of ocean crust. Since there is only continental thrust over 200 km (a hypothesis that G.D. feels is mechani- crust to the north, then a south-dipping subduction zone cally unreasonable). The tectonic provenance and metamor- seems an unlikely explanation. Equally there appears to be phism of the Purial Complex also remains unresolved in this no evidence of a Paleogene, north-dipping subduction zone model. (major accetionary complex or blueschist complex) south of 2. In the second scenario (Fig. 4.15), the arc was south- Cuba or in Hispaniola. The only possible candidate might facing and a small ocean basin separated it from the passive be the Trois Rivieres-Peralta belt of Hispaniola (see Draper 30 margin (Iturralde-Vinent, personal communication). The et al., Chapter 7, this volume). collision event would have resulted when buoyant (unsub- The plate kinematic cause of the early Eocene (second ductable) ocean crust, represented by the Escambray and phase) deformation in central Cuba also remains obscure, Isla de la Juventud blocks, entered the subduction zone. As but may have been associated with the movement of the a result the Escambray (and Purial?) rocks were metamor Caribbean plate into the meso-American region and the phosed to blueschist facies under the fore-arc, the ocean opening of the Yucatan basin (although this was an exten- basin behind the arc was thrust onto the passive margin, and sional, not a compressional, event). subduction ceased. In this scenario the Cuban ophiolites represent fragments of the basement beneath the arc and/or Mid-Eocene to Recent the small, marginal bas in, but not the major Proto-Caribbean From the middle Eocene to the present, much of Cuba

81 Cuba

has been (relatively) tectonically quiescent. By examining 6Draper, G. 1989b. Comparison of late Cretaceous -Paleo- later Paleogene and Neogene sedimentary deposits, Ittu- gene age rocks of northern Hispaniola and southeastern 14,15 ralde-Vinent determined that post-orogenic Cuba con- Cuba: implications for the tectonic evolution of the sists of a series of differentially subsiding horsts and grabens Greater Antilles. Primer Congreso Cubano de Geolo- (Fig. 4.16). Subsidence was most rapid and widespread gia, Resumenes y Programa, La Habana, 211. during the late Eocene and Oligocene, but continued 7Draper, G. & Barros, J.A. 1988. Tectonic reconstruction of through the Miocene and Pliocene in several areas. Pleisto- N. Hispaniolaand S.E. Cuba: dissection of a Cretaceous cene to Recent sedimentation was of moderate extent. island arc. Geological Society of America Abstracts The major, post-orogenic tectonic activity was, and is, with Programs, 20 (7), A60. south of Oriente on the Oriente transform fault system that 8Ducloz C. & Vaugnat, M. 1962. A propos de 1'age des separates Cuba from the present Caribbean plate. As men- serpenitites de Cuba. Archives de Science Societe de tioned above, several authors consider that southeastern Physique et Histoire Naturell de Geneve, 15, 309-332. Cuba and northern Hispaniola were once continuous, al- 9Furrazola-Bermudez, G. et al. 1964. Geologia de Cuba. though the precise palaeogeography is still uncertain. It is Ministerio de Industrias, Institute Cubano Recursos also unclear as to when this separation began. Extensive Minerales, La Habana, 239 pp. clastic sedimentation, probably related to rapid uplift and 10Gealey, W. K. 1980. Ophiolite obduction mechanisms: in erosion, began in the middle to late Eocene in both south- Panayiotou, A. (ed.), Ophiolites: Proceedings, Interna- eastern Cuba and northern Hispaniola. We suggest here that tional Ophiolite Symposium, Nicosia, Cyprus, 1979, this was the time of initiation of the separation of the two 228-243. Geological Survey Department, Cyprus. areas. Extrapolation of the spreading rates in the Cayman 11Hatten, C., Schooler, C.H., Giedt, N.R. & Meyerhoff, Trough31 also suggests that major motion on the Cayman 5-7 A. A. 1958. Geology of central Cuba and western Cam- transform fault also began at this time . Motion continues aguey provinces. Unpublished Report, Chevron on this fault and the region is seismically active. U.S.A., 220 pp. 12Hatten, C., Somin, M., Millan, G., Renne, P.R., Kistler, ACKNOWLEDGEMENTS—We thank all of our Cuban and eastern Euro- pean colleagues who have taught us much about the geology of Cuba. In R.W. & Mattinson, J.M. 1987. Tectonostratigraphic particular, we would like to thank Jose Francisco Albear, Karoli Brezsny- units of central Cuba: in Barker, L. (ed.), ansky, Mario Campos, Jorge Cobiella, Jose Diaz-Duque, Gustavo Echavar- Transactions of the Eleventh Caribbean Geological ria, Rafael Guardada, Margarita Hernandez, Manuel Itturalde-Vinent, Conference, Dover Beach, Barbados, 20th-25th Guillermo Millan, Yurek and Kristina Piotrowska, Andre Pszczolkowski July, 1986, 38:1-38:13. and Felix Quintas, all of whom have corresponded copiously, showed us 13 outcrops in the field and generously shared unpublished data and reports. Herrara, N.M. 1961. Contribution a la estratigrafia de la Outside of Cuba, animated discussions with Charles Hatten, Mark Hemp- Provincia de Pinar del Rio. Revista Cubana de la So- ton, John Lewis, Paul Mann, Paul Renne and Eric Rosencrantz have also ciedad de Ingenieros, 61, 2-24. helped stimulate our interest in, and clarify ideas about, Cuban geology. 14Itturalde-Vinent, M. 1977. Los movimientos tectonicos de la etapa de desarollo plataformico en Cuba. Academia de Ciencias de la Tierra, Informe REFERENCES Cientifico-Tecnico, 20, 24pp. 15Itturalde-Vinent, M. 1978. Los movimientos tectonicos de 1Barros, J.A. 1987. Stratigraphy, structure and paleo- la etapa de desarollo plataformico en Cuba. Geologie geography of the Jurassic-Cretaceous passive margin en Mijnbouw, 57, 205-212. in western and central Cuba. Unpublished M.S. thesis, 16Khudoley, K.M. 1967. Principal features of Cuban geol- University of Miami, Florida. ogy. American Association of Petroleum Geologists 2Cobiella, J.L. 1984a Curso degeologia de Cuba. Editoral Bulletin, 51, 789-791. Pueblo y Education, La Habana, 114 pp. 17Lewis, J.F. 1990. Cuba [in Lewis, J.F. & Draper, G. 3Cobiella, J.L. 1984b. Position de Cuba Oriental en la Geological and tectonic evolution of the northern Car - geologia del Caribe. Revista Mineria y Geologia, 2, ibbean margin]: in Dengo, G. & Case, J.E. (eds), The 65-92. geology of North America, Volume H, The Caribbean 4Dilla, M. & Garcia, L. 1985. Nuevos datos sobre la estar - region, 77-140. Geological Society of America, Boul- tigrafia de las provincias de Cienfuegos, Villa Clara y der. Sancti. Espiritus serie Geologica, 5, 54-77. 18Mattson, P. 1974. Cuba: in Spencer, A.M. (ed.), Meso- 5Draper, G. 1989a. Terrane accretion and re-shuffling in zoic-Cenozoic Orogenic Belts. Geological Society of Hispaniola. Geological Society of America Abstracts London Special Publication, 4, 625-638. with Programs, 21(1), 9. 19Meyerhoff, A. A. & Hatten, C. 1968. Diapiric structures in

82 G. DRAPER and J. A. BARROS

central Cuba. American Association of Petroleum phy. Acta Geologica Polonica, 28, 1-96. Geologists Memoir, 8, 315-357. 28Pszczolkowski, A. & Flores, R. 1985. Fases tectonicas del 20Millan, G. 1981. Geologia del macizo metamorfico de la Cretacicos y del Paleogeno de Cuba occidental y cen- Isla de la Juventud. Ciencias de la Tierray Espacio, 3, tral. Serie Geologica, 20. 5-22. 29Renne, P.R. et al. 1989. 40Ar/39Ar and U-Pb evidence 21Millan, G. & Somin, M. 1976. Agunas consideraciones for late Proterozoic (Grenville age) continental sobre la metamorfitas cubanas. Academia de crust in north-central Cuba and regional tectonic Ciencias de Cuba, serie geologica, 27, 21 pp. implications. Precambrian Research, 42, 325-341. 30 22Millan, G. & Somin, M. 1981. Litologia, estratigrafta, Rosencrantz, E. & Barros, J.A. 1989. Structural and tectonica y metamorfismo del macizo del Escambray. chronological discontinuity in late Cretaceous- Academia de Cienias de Cuba, Habana, 104 pp. Eocene tectonic events in Cuba. Geological Society of 23Mossakovsky, A. & Albear, J.F. 1978. Nappe structure of America Abstracts with Programs, 21(1), 39. 31 western and northern Cuba and its history in the Rosencrantz, E. & Sclater, J.G. 1986. Depth and age in the light of the study of olistostromes and mollases. Cayman Trough. Earth and Planetary Science Letters, Geotectonics, 12, 225-236. 79, 133-144. 32 24Mossakovsky, A. et al. 1988. Mapa geologico de Somin, M.L. & Millan, G. 1977. Sobre la edad de las rocas Cuba,escala 1:250,000. Academia de Ciencias de metamorficas Cubanas. Informe Cientifico-Tecnico, Cuba, editado por el institute de Geologia de Ciencias 2, llpp. de URSS H-2111. 33Somin, M.L. & Millan, G. 1981. Geology of the metamor- 25Pardo, G. 1954. Geologic exploration in central Cuba. phic complexes of Cuba. Nauka, Moscow, 218 pp. Unpublished Report, Cuban Gulf, 79 pp. 34Wassal, H. 1956. The relationship of oil and serpentinites 26Pardo, G. 1975. Geology of Cuba: in Nairn, A.E.M. & in Cuba: in Transactions of the 20th International Stehli, F.G. (eds), The Ocean Basins and Margins. Geological Congress, 65-77. Volume 3. The Gulf of Mexico and the Caribbean, 35Weyl, R. 1966. Geologic der Antillen: in Martini, H.J. 553-615. Plenum, New York. (ed.), Band 4, Beitrage zur Regionalen Geologie 27Pzczowlkowski, A. 1978. Geological sequences of the der Erde: 1-410. Gebriider Bomtrager, Berlin. Cordillera de Guaniguanico in western Cuba: their li- thostratigraphy,facies development and paleogeogra-

Cuba

APPENDIX 1

Map symbols used in Cuban literature Many maps, sections, columns and diagrams of Cuban geology use a different notation for the ages of rocks than is otherwise used in the literature of the Caribbean and North and South America. The notation principally derives from a system developed in eastern Europe. It is presented here for those readers who may wish to delve deeper into the

Cuban literature. System Series Stage

Quaternary (Q) Holocene (Q2)

Pleistocene (Q1)

Neogene (N) Pliocene (N2)

Miocene (N1)

Paleogene (P) Oligocene (P3)

Eocene (P2)

Paleocene (P2)

Cretaceous (Upper) (K2) Maastrichtian (K2m)

Campanian (K2cp)

Santonian (K2st)

Coniacian (K2cn)

Turonian (K2t)

Cenomanian (K2c)

Albian(K1al)

Cretaceous (Lower) (K1) Aptian (K1ap)

Barremian (K1bm)

Hauterivian (K1h)

Valanginian (K2v)

Berrasian (K1b)

Jurassic (Upper) (J3) Tithonian (J3t)

Kimmeridgian (J3k)

Oxfordian (J30x)

Jurassic (Middle) (J2) Callovian (J2ca)

Bathonian (J2bt)

Bajocian (J2bj)

Series and stages can be further subdivided by adding superscripts to the symbols. 1 For example, Lower, Middle and Upper Eocene rocks would be denoted as P2 , 2 3 P2 and P2 , respectively.

G. DRAPER and J. A. BARROS

APPENDIX 2 Sub-Upper Eocene formations in the structural facies zones of the four main structural blocks of Cuba.

85 Cuba

CHAPTER 5

The Cayman Islands

BRIAN JONES

Department of Geology, University of Alberta, Edmonton, Alberta, Canada T6G 2E3

INTRODUCTION or more3,13. Perfit and Heezen65 suggested that the present day Cayman Trench was initiated in Eocene times. The GRAND CAYMAN, Cayman Brae, and Little Cayman, Cayman Ridge was predominantly a shallow water carbon- collectively known as the Cayman Islands, are located on ate bank until Miocene times when it began to subside65. As the Cayman Ridge in the central Caribbean Sea (Figs. 5.1, subsidence continued, at 60 to 100 mm per 1000 years12,65, 5.2). Matley55 named the Bluff Limestone for the hard, areas of shallow water carbonate accumulation and reef resistant carbonates that formed the core of each island and development were progressively restricted to the higher the Ironshore Formation for the overlying softer limestones parts of the ridge. Following middle Miocene times, local- (Fig. 5.3). This stratigraphic scheme is deceptively simple ized uplift elevated Central America and raised Swan Island, because it embodies subtle complexities which denote im- the Cayman Islands, Jamaica and most of southern Cuba portant events in the geological evolution of the islands. An above sea level65. The fact that Oligocene-Miocene carbon- understanding of the evolution of the Cayman Islands can ates of the Cayman Islands are now exposed above sea-level only be obtained by a fully integrated investigation of their clearly records the change from subsidence and deposition geology. to uplift and erosion. The differential uplift of these islands led to the suggestion that the Cayman Ridge is divided into discrete blocks by northeast-southwest trending faults, each TECTONIC SETTING OF THE CAYMAN island being on a separate block55,76. Horsfield20 pointed out ISLANDS that there is a general tectonic tilt to the west such that the islands become progressively lower in that direction. The three Cayman Islands are situated on the Cayman Ridge The three Cayman Islands are pinnacles (Fig. 5.2) rising (Figs. 5.1, 5.2) which forms the northern margin of the up from the depths of the ocean. Each island appears to be Cayman Trench that is 100-150 km wide51, 81 and reaches 51 based on a foundation of granodiorite which occurs at depths depths in excess of 6000 m . The Mid-Cayman Rise is between 600 and 3400 m on the Oriente Slope to the south located to the southwest of Grand Cayman at about 82° W of Grand Cayman12,76. The granodiorite is succeeded by a (Fig. 5.1). The Oriente Transform Fault forms the northern cap of basalt76 and then by Tertiary carbonates. The thick- boundary of the Cayman Trench to the east of the rise ness of the carbonate cap on Grand Cayman is not known. whereas the Swan Island Transform Fault forms the south- Two wells drilled in the central part of the island, however, ern boundary of the Cayman Trench to the west (Fig. 5.1). were still penetrating Oligocene carbonates at depths of 401 These transform faults are characterized by the left-lateral m (1315 feet) and 159 m (520 feet), respectively, when the motion of the North America Plate relative to the Caribbean wells were capped Plate19,51,65. Seismic activity along the Mid-Cavman Rise (Fig. 5.1) and adjoining transform faults74,78,79 indicates that the spreading centre is still active51. THE BLUFF FORMATION The Cayman Ridge appears to be an uplifted fault block10,13,52,65 that rises 1500-2000 m above the surrounding Introduction seafloor and is bounded by fault planes dipping at 30° Matley55 originally named the Bluff Limestone for

87 The Cayman Islands

Figure 5.1. Map of central and northern part of the Caribbean Sea showing the positions of the Cayman Islands relative to the major tectonic features of the area.

Figure 5.2. East-west and north-south cross-sections across the Caribbean Sea showing the general setting of the Cayman Islands.

poorly-bedded, white to cream coloured, hard, microcrys- Castle (Fig. 5.3 A) as the type section (Fig. 5.5). The Cayman talline, porous carbonates that formed the core of each island Member is separated from the overlying Pedro Castle Mem- (Fig. 5.3). The idea that this formation was formed entirely ber by a disconformity (Fig. 5.5). of limestone persisted 2,5,6,53,54,55,71,15,76,89,92,93 until it was demonstrated 39,48,66,68 that it is formed almost entirely of The Cayman Member dolostone. Consequently, Jones and Hunter33 renamed this Boundaries unit as the Bluff Formation (Fig. 5.4) in order to overcome The lower boundary of the Cayman Member is not the problems caused by the lithological connotation attached exposed on any of the Cayman Islands. The upper boundary to the term Bluff Limestone. is placed at the disconformity which is well-exposed in the Jones and Hunter33 divided the Bluff Formation into quarry near Pedro Castle (Fig. 5.5). The type section, as the Cayman and the Pedro Castle Members, and designated presently defined, only covers the upper 5 m of the member. a succession in the northwest corner of the quarry near Pedro 88 BRIAN JONES

Lithology impossible to obtain good sections through the formation. At its type section, Jones and Hunter33 divided the On Cayman Brae (Fig. 5.3C) the Bluff Formation ap- Cayman Member into units I and II (Fig. 5.5). Unit I (3 m pears to be formed entirely of the Cayman Member. The thick) is a white, hard, microcrystalline dolostone that was Cayman Member forms vertical cliffs that rise up to 50 m originally a packstone to grainstone. Scattered colonial above sea level at the northeast end of the island. Indeed, the (Montastrea, Leptoceris, Agathiphyllia, Favia, Diploria, term Bluff was originally used by Matley55 to denote the and Diploastrea crassolameletta) and branching (Stylo- bluff-forming characteristics of the formation on Cayman phora and Porites) corals, gastropods (including conch Brac. Although the cliff faces are largely inaccessible, good shells), and rare bivalves occur throughout the unit. Most of exposures of the formation can be seen along road cuts and in these fossils are now represented by moulds because the quarries. original aragonite was removed by leaching. Moulds of the corals, gastropods and bivalves commonly contain casts of The Pedro Castle Member borings (such as Entobia, Trypanites, Uniglobites)66,67. Boundaries Skeletal components (foraminifers, red algae fragments, The lower boundary of this member is placed at the echinoid plates and spines) originally formed of high mag- disconformity which is exposed in the quarry near Pedro nesium calcite are well-preserved. Unit I is characterized by Castle (Fig. 5.5). The upper boundary is the unconformity high porosity (up to 25%) and permeability. All cavities are with the overlying Ironshore Formation. In most areas the lined with two generations of dolomite cement39. Some present-day erosion surface cuts down into the Pedro Castle cavities are filled or partly filled with caymanite (white, red, Member. and black laminated dolostone29,50,71). Lithology Unit II (1 .75 m thick) is a white, microcrystalline dolos- This member is 2.5 m thick at the type section (Fig. 5.5). tone that contains abundant branching corals (Stylophora It is formed of off-white to cream coloured, relatively soft and Porites), bivalves, gastropods, rare complete echinoids, (compared to the Cayman Member), rubbly-weathering rare colonial corals, and foraminifers33. The fossils are dolostone that contains numerous large free-living corals embedded in a mudstone to wackestone (Fig. 5.5). As in unit (including Trachyphyllia, Thysanus, Teliophyltia and Antil- I, the corals, bivalves, and gastropods have been leached, and locyathus), abundant well-preserved foraminifers (mainly are now represented by moulds. Many cavities are partly or miliolids), rare fragments of red algae, rare colonial corals, completely filled with white caymanite and, more rarely, red echinoid fragments, and bivalves. Rhodolites, up to 5 cm in and black caymanite. diameter, are common in the basal bed of the member. As in Age the Cayman Member, the corals, gastropods, and bivalves are Jones and Hunter33 suggested that the fauna in the represented by moulds because the aragonite has been Cayman Member exposed on Grand Cayman was of early removed by leaching. late Oligocene age. This is in general agreement with the Some cavities in the Pedro Castle Member are filled or late Oligocene age that Vaughan82 suggested for equivalent partly filled with unconsolidated terra rossa that has been strata on Cayman Brae. This age is for the upper beds of the derived from the overlying soil (Fig. 5.5). Unlike the Cay- member because only the upper part of the member is man Member, the pores and cavities are not lined with exposed on the islands. The age of the lower part of the dolomite cements. White caymanite and dolomitized member is not known because it is not exposed on Grand foraminiferal grainstones are rare in the cavities of this Cayman. member. Distribution Age On Grand Cayman, the Cayman Member crops out on Determining the age of the Pedro Castle Member is the eastern part of the island where it is exposed in outcrops difficult because of the lack of biostratigraphically significant up to 17 m above sea level (Fig. 5.3 A). The formation also fossils. Jones and Hunter33 suggested, however, that the forms a narrow ridge on the south and north coasts of the corals were probably indicative of a Miocene age (pos sibly island (Fig. 5.3A). Isolated outcrops of this member occur to Middle Miocene). the south of George Town and at Hell (Fig. 5.3 A). Distribution On Little Cayman (Fig. 5.3B) the exposed strata belong The Pedro Castle Member is restricted to the area around to the Cayman Member of the Bluff Formation. As with its Pedro Castle on Grand Cayman (Fig. 5.3A). It has not been sister islands, the Bluff Formation forms the core of the identified on Little Cayman or Cayman Brae. island. Like Grand Cayman, Little Cayman is characterized On Grand Cayman, it is well-exposed in the quarry near by a subdued topography with maximum elevations no more Pedro Castle, and along the coastline between Pedro Castle than 12 m above sea level. The exposures are poor and it is and Spots Bay (Fig. 5.3A). Although present in the area

89 The Cayman Islands

Figure 5.3. Maps of (A) Grand Cayman, (B) Little Cayman and (C) Cayman Brae, showing the distribution of the Bluff and Ironshore Formations. around Savannah, it can only be seen when house founda- white to cream, in colour; (iii) its cavities contain a wide tions are being dug or in small pinnacles that occur in some of variety of cements rather than no cements; and (iv) it con- the swamp areas to the north. The Pedro Castle Member was tains branching and massive corals rather than large, free- present in the northeast corner of the airport (locality ARP in living corals33. Fig. 5.3A) prior to its removal during excavations associated with runway extension. Contact between the Cayman and Pedro Castle members At the type section, the contact between the Cayman Comparison of Cayman and Pedro Castle Members and Pedro Castle members is marked by adisconformity that At its type section, the Cayman Member is distin- dips at about 2° to the northwest33. Locally, the disconfor- guished from the Pedro Castle Member (Figs. 5.5,5.6,5.7) mity is encrusted with red algae. The upper surface of the because (i) it is massive and hard, rather than rubbly and Cayman Member was extensively bored by sponges, relatively soft; (ii) it is off-white to light grey, rather than worms, and bivalves prior to deposition of the Pedro Castle

90 BRIAN JONES

Member. These borings indicate that the Cayman Member Bivalve facies must have been lithified prior to deposition of the Pedro This mudstone to wackestone contains a well-pre- Castle Member. served, abundant, diverse mollusc fauna8,9,69,70, that is developed in the eastern part of the Ironshore Lagoon. The molluscan fauna, which includes 83 species of bivalves and THE IRONSHORE FORMATION 90 species of gastropods, is dominated by burrowing bi- valves8,9. Corals are rare (Table 5.1). Locally, foraminifera Introduction and fragments of red algae are common. Hunter and 55 Matley defined the Ironshore Formation and noted Jones 21,22,34 suggested that this facies developed in shallow that it contained abundant well-preserved corals. Corals water in the landward portion of the Ironshore Lagoon. The from the upper part of the formation have yielded ages of nature of the mollusc fauna led Jones and Hunter34 to 93 124,000±8000 years . The formation was not described in suggest that the muddy seafloor was covered with Thalas- 5 detail until Brunt et al divided it into (1) reef, (2) back-reef, sia. (3) lagoonal, (4) oolitic, and (5) shoal facies. Woodroffe et 92 al. subsequently examined the lagoonal facies in the Coral A facies northwest part of Grand Cayman and demonstrated that it This floatstone is dominated by a well-preserved, abun- encompassed a series of patch reefs surrounded by lagoonal 44,45 dant and diverse coral fauna (Table 5.2) along with lesser deposits. Jones and Pemberton demonstrated that many numbers of bivalves and gastropods (Table 5.1). This facies corals from the patch reefs in the Ironshore Formation were 21,22,31,46,64 occurs in rounded to elliptical areas that are generally less extensively bored by Lithophaga.. Recent work than 100 m in diameter (maximum 200 m long). Dominant has shown that the formation contains a complex array of corals include Montastrea annularis, M. cavernosa, facies that varies both laterally and vertically. 34 Diploria strigosa, D. labyrinthiformis, and Siderastrea Jones and Hunter showed that the upper part of the siderea (Table 5.2). Associated with the corals are numerous Ironshore Formation records a shallowing-upward se- bivalves that either nestled between the corals or cemented quence. The limestones in this part of the formation were themselves to the solid substrates8,9. Many corals in these deposited in a complex array of depositional environments 21,34 areas suffered considerable bioerosion through the activity in a large lagoon which Hunter and Jones called the of boring bivalves44, sponges, worms, algae, and fungi22,34. Ironshore Lagoon (Fig. 5.8). The aerial restriction of the corals and the nature of the fauna has led to the suggestion 21,34,92 that they represented patch

Palaeogeographic framework reefs. Corals in these patch reefs were capable of sediment The palaeogeographic framework of the Ironshore For- rejection21,34. mation was largely controlled by the rugged karst topography that developed on the Bluff Formation during middle Coral B facies 22,34 Miocene to late Pleistocene times . Although the main This floatstone contains an abundant, diverse coral area of deposition was in the Ironshore Lagoon that covered fauna (Table 5.2) with many corals still in growth position. most of the western half of Grand Cayman, there were small It is separated from coral facies A because it is not aerially embayments along the south, east, and north coasts which restricted into patch reefs. In addition to the corals that occur were akin to the sounds that occur around the modem 21,34 in coral facies A, this facies also contains abundant shoreline. According to Hunter and Jones , the north and Acro-pora palmata, A. cervicornis, Dendrogyra cyclindrus south margins of the Ironshore Lagoon were delineated by and Podllopora sp. Unlike the corals in the Coral A facies, narrow ridges of the Bluff Formation (Fig. 5.7B). The these corals are not bored. This facies occurs on the western margin of the lagoon was marked by a barrier reef northwest, west and southwest coasts of Grand Cayman (Fig. 5.8C). Facies in the Ironshore Formation have been 5,21,22,31,46,64,92 (Fig. 5.8). The composition of the coral fauna is highly defined in many different ways because dif- variable between localities (Hunter, 1990, personal ferent studies have emphasized different aspects of the communication). rocks. Many schemes named the facies with respect to Interpretation of the coral communities suggests that their depositional setting and thus relied heavily on they lived in a back-reef setting rather than on the reef interpretation. Such problems are avoided by usin g a non- crest22,34. In contrast to the patch reefs, many corals in this genetic classification based on lithology, sedimentary facies could not reject sediment. structures, fossil content and ichnology23,34. The scheme outlined herein (Table 5.1) is a combination of the 23,34 Laminated to highly burrowed grainstone facies schemes used by Hunter and Jones . This facies (Table 5.1) includes both skeletal and ooid grainstones, and occurs between the patch reefs. The skele- 91 The Cayman Islands

Figure 5.4. Schematic stratigraphic section for the Cayman Islands. Inset shows stratigraphy of internal sediments which fill cavities in the Cayman Member of the Bluff Formation. tal grainstones commonly contain a diverse bivalve Unidirectional, high-angle cross-bedded grainstone facies fauna8,9,92, whereas the oolitic grainstones contain only a 46,64 This facies (Table 5.1), which only occurs at locality few bivalves. Burrowing (Table 5.1) destroyed most SC (Fig. 5.8A), overlies an erosional surface that has large, original sedimentary structures. The occurrence of this fa- tabular lithoclasts resting on it. Woodroffe91 attributed a cies between the patch reefs suggests that it accumulated in a subaerial origin to this facies. However, Jones and co-work- shallow, lagoonal setting. ers34,46,64 argued that it originated in subtidal conditions because it is restricted to a channel and contains an ichno-

Rudstone facies fossil assemblage that could only have developed in a sub- This facies (Table 5.1), that only occurs south of Salt tidal environment. Creek on the west coast of North Sound (locality SC in Fig. 5.8A), is formed of lithoclasts (up to 0.1 m x 0.1 m x 0.1 m) Laminated to low angle cross-bedded grainstone facies of oolitic grainstone cemented by equant spar calcite, or This facies (Table 5.1) is best exposed at localities BQ, bioclasts (corals and bivalves) in an oolitic grainstone ma- LSC, PBQ, PBA, and SBA (Fig. 5.8A). Interpretation of the trix. This facies, which rests on an erosional surface, records 34,46 lithotypes, sedimentary structures, and ichnofossils led a high energy event such as a storm or hurricane . Jones and Hunter34 to suggest that it recorded deposition in a swash zone. This succession is comparable to similar facies Multidirectional, high-angle cross-bedded grainstone in the Pleistocene of the Yucatan Peninsula which are facies considered indicative of that zone84,85. This facies (Table 5.1) is well exposed at localities BQ, LSC, SC, PBQ, PBA, SBA, and ACH (Fig. 5.8A). Near Sea level during deposition of the Ironshore Formation localities SC and LSC this facies overlies an erosional 91 46 Although Woodroffe argued that sea level was only surface on which tabular lithoclasts (up to 0.3 m x 0.3 m x 2 to 3 m above present-day sea level at the time when the 0.1 m) rest. The presence of Ophiomorpha and Conichnus Ironshore Formation was deposited, Jones and Hunter 34 shows that this facies accumulated in subtidal conditions64. 34 demonstrated that it was actually at about 6 m. This estimate of Jones and Hunter suggested that it represents an upper the sea level is in agreement with that generally given for this shoreface (surf zone) setting. age of deposits throughout the Caribbean4,7,1 7,56,60.

92 BRIAN JONES

Vertical facies succession On Little Cayman there are few exposures through the 76 It is impossible to determine the vertical facies succes- Ironshore Formation. Stoddart briefly described and illus- sion for the entire thickness of the Ironshore Formation trated a sequence near Salt Rocks that is akin to that on because most of it is below sea level. Consequently, the Grand Cayman because it encompasses a shallowing-up- 34 vertical facies succession documented herein focuses on the ward succession . A quarry on the south-central coast upper 7 m of the formation that is exposed at various demonstrates that the Ironshore Formation is capped by a localities on the western part of Grand Cayman (Fig. 5.9). thick caliche unit similar to that on Cayman Brae. The upper part of the Ironshore Formation on Grand Cayman records a shallowing-upward succession (Fig. 5.9). BOUNDARY BETWEEN THE BLUFF AND The lower part of the sequence, which formed in the Iron- IRONSHORE FORMATIONS shore Lagoon, ranges from the lagoonal deposits (bivalve facies) in the east to the patch reefs (coral A facies), inter- Determining the profile of unconformity between the Bluff reef grainstones (laminated to highly burrowed grain- and Ironshore Formations is difficult because much of it is stones), and reef tract (coral B facies) deposits in the west below sea level and buried by the Ironshore Formation. (Fig. 5.8). Locally, as near Salt Creek, these lagoonal depos- Nevertheless, the sparse information that is available sug- its were cut by large tidal channels34,46. Some channels have gests that the unconformity has considerable relief on it. rudstones or isolated clasts resting on their floors34,46. This is amply demonstrated by the fact that isolated patches As sea level dropped, the lagoon became progressively of the Bluff Formation protrude through the Ironshore For- shallower and a complex array of facies formed in response. mation at Hell, south of George Town and at various loca- Thus, overlying the lagoonal deposits are a succession of tions along the north and south coasts (Fig. 5.3A). This is cross-bedded grainstones and laminated grainstones that also shown by the fact that the unconformity is essentially record the transition from the lagoon through the lower horizontal at Newlands, but vertical at Spots Bay (locality shoreface to the upper shoreface34. Valuable insights into SBJ in Fig.5.8A). the nature of the last stages of the regression were obtained from an analysis of a community replacement sequence35. This sequence involved the death of corals, boring of the DIAGENETIC FABRICS IN THE BLUFF coral fragments, coating of the coral fragments by red algae FORMATION and foraminifers to form rhodolites, and, finally, binding of 35 The Bluff Formation includes a vast array of diagenetic micrite by microbes to form microbialites . fabrics that reflect the complex history of this formation The Ironshore Formation on Cayman Brae and Little since middle Miocene times. Despite its relatively young Cayman age, the rocks of this formation have been subjected to The stratigraphy and sedimentology of the Ironshore pervasive dolomitization, numerous cycles of karst devel- Formation on Cayman Brae and Little Cayman is poorly opment, numerous phases of dissolution, cement precipita- known because few studies have been done on the relatively tion and deposition of internal sediment. Locally, diagenesis poor exposures through the formation on those islands. has obscured the original depositional fabrics. Porosity and On Cayman Brae the limestones of the Ironshore For- permeability is generally high in the Bluff Formation because of the vast array of diagenetic processes that have mation appear to have accumulated in narrow lagoons that 63 fringed steep cliffs of the Bluff Formation. A well-devel- affected the rocks . An understanding of the diagenetia oped, wave-cut notch at about 6 m above present day sea fabrics is important because they are the only record of the level34,93 delineates the position of the late Pleistocene conditions and processes that occurred following deposi- sea level. Near Pollard Bay, the lagoonal deposits are tion. characterized by numerous corals that are still in life position. It appears that coastal erosion was active during Dolomitization that time because there is evidence of sea-cliff collapse Pervasive dolomitization is the most obvious change caused by undercutting of the dolostone cliffs42. that the rocks of the Bluff Formation underwent. Despite Small pits near the airport on Cayman Brae cut through dolomitization, the original fabrics of the rocks were largely the upper 5 m of the Ironshore Formation and show a retained and the nature of the original sediments can still be sequence which records deposition from the lower to upper determined. For example, fossils originally formed of high- shoreface43. The succession is capped by a well-developed magnesium calcite (for example, red algae, foraminifers and caliche unit that has been extensively bored by a wide echinoids) have retained their original skeletal structure. variety of plant roots26. Despite extensive study, the conditions responsible for

93 The Cayman Islands

Figure 5.5. Summary of depositional and diagenetic features in the type section of the Bluff Formation at Pedro Castle Quarry (modified after Jones and Hunter33; Jones et al.36).

the dolomitization of the Bluff Formation are not fully of dolomitization reset the chemical signature of the rocks understood. The geographic isolation of the Cayman Islands in the Cayman Member and thus gives the false impression as well as the geological framework of the Bluff Formation of a uniform signature. Although a viable possibility, there means that many of the standard dolomitization models are no textures or geochemical signatures that support the cannot be used to explain the dolomitization. Thus, Pley- contention of more than one phase of dolomitization. Inter- dell66 and Pleydell et al69 suggested that dolomitization of pretation of the 87Sr/86Sr ratios from the dolostones led the Bluff Formation was mediated by 'normal' seawater or Pleydell et al.68 to suggest that dolomitization of the Bluff by mixed seawater and freshwater. Formation was probably caused by 'normal' seawater 2 to The dolostones in the Cayman and Pedro Castle Mem- 5 million years ago. The reasons why dolomitization oc- bers are petrographically and geochemically similar. For curred during that time are not yet understood. example, the dolostones of these two members have similar C, O, and Sr isotope signatures68. These similarities have Dedolomitization led to the suggestion that the Bluff Formation had only In some parts of the Bluff Formation there are soft, undergone one phase of dolomitization. However, Ng62 porous, 'chalky' zones that contrast sharply with the hard argued that the Bluff Formation had been subjected to two dolostones which usually characterize the Bluff Formation. phases of dolomitization, the first following deposition of Such zones occur along joints, fractures and bedding planes the Cayman Member, but before the deposition of the Pedro which allow easy passage of various waters47,61. These Castle Member, and the second following deposition of the rocks are formed of hollow dolomite rhombs47 which re- 62 Pedro Castle Member. According to Ng , the second phase sulted from the cores of dolomite crystals eing preferen- 94 BRIAN JONES

Figure 5.6. Comparison of depositional features and fauna in the Cayman and Pedro Castle Members of the Bluff Formation at its type section (after Jones et al36).

Figure 5.7. Comparison of diagenetic features in the Cayman and Pedro Castle Members of the Bluff Formation at its type section (after Jones et al.36). tially dissolved. If present, the calcite cement around the showed that there are three main generations of cements in dolomite rhombs forms a poikiloptopic texture47. The de- the dolostones of the Cayman Member. The first generation dolomitization appears to be a relatively recent phenomenon cement, which occurs in most cavities in the Cayman Mem- that may be related to modern groundwater activity. ber, is formed of euhedral crystals of zoned or unzoned limpid dolomite. Jones et al.39 and Ng62 suggested that this Cements cement formed from mixed freshwater and saltwater in the Many cavities in the Cayman Member are partly or phreatic zone. The second generation cement, formed of totally occluded by a complex array of cements that formed coarsely crystalline spar calcite, probably originated in the under a wide variety of conditions 25,37,38,39,62. Jones et al.39 freshwater phreatic zone39,62. The third generation cement,

95 The Cayman Islands

formed of microstalactites with alternating bands of calcite surface successions that have since been removed by ero- and dolomite, developed while the rocks were in the vadose sion. Some cavities and sinkholes in the Bluff Formation are zone. The three generations of cement record the passage of partly or completely filled with terrestrial oncoids 28. These the rocks in the Cayman Member through the mixed-water ovate, coated grains, which are generally less than 5 mm phreatic, freshwater phreatic, and vadose zones. Cavities in long, formed through the action of a diverse array of micro- the dolostones of the Pedro Castle Member are not lined organisms. with limpid dolomite33. This suggests that the limpid dolomites in the Cayman Member formed before deposition of Joints the Pedro Castle Member. Locally, cavities in the Pedro In many areas the Bluff Formation is transected by Castle Member are partly or totally occluded by spar calcite. well-developed joints48,71. Ng62 suggested that lineaments visible on air photographs are surface expressions of these Internal sediments joint sets. The joints are important because they act as Many caves and cavities in the Bluff Formation are conduits by which seawater penetrates into the central part filled or partly filled with caymanite, dolomitized foraminif- of the island. Many joints are partly or totally filled with a eral grainstone, unlithified and lithified terra rossa, breccia, complex array of precipitates and internal sediments48. For and/or terrestrial oncoids. example, in the area near Blowholes, the joints are filled Caymanite is a red, black, and white (or any shade with caymanite, terra rossa and terra rossa breccia48. Lo- 48 between these colours), laminated, cryptocrystalline dolos- cally, the walls of the joints are lined with flowstone . The tone that derives its name from the fact that the local people fact that the flowstone and terra rossa are not dolomitized use it for making jewellery71. Lockhart50 demonstrated that suggests that these fills, and possibly the joints, were formed the colour was due to trace amounts of Fe (red) and Mn after dolomitization. (black). Folk and McBride14 and Lockhart50 suggested that the coloured caymanite was formed when transgressing seas Karst development washed swamp deposits through the karst system into the The present day surface of the Bluff Formation is caves and caverns. Jones 29 demonstrated that caymanite characterized by a rugged karst terrain that includes jagged 11,15,27,83 was formed by sediments from numerous sources (marine, surfaces, potholes and solution-widened joints . ponds, swamps and soils) being washed into the subterra- Indeed, the exposure of the Bluff Formation at Hell repre- nean cavities by water draining from the surface of the karst sents the type locality of phytokarst as described by Folk et 15 terrain. The colours and sedimentary structures which char- al. . In addition to the surface features, there is a well-de- acterize caymanite formed in response to a complex set of veloped subterranean karst system with numerous caves that interrelated variables29. commonly contain a diverse array of stalactites, stalagmites, 40,41,48 Some cavities and caves in the Cayman Member ex- and various other speleothems . posed in the quarry near Pedro Castle contain a dolomitized The Bluff Formation has undergone numerous phases grainstone that has high mouldic porosity. The presence of of karst development. Study of outcrops in the quarry near foraminifers and fragments of red algae indicates that the Pedro Castle over the last ten years has yielded important sediment was of marine origin. information concerning karst development relative to the Many cavities in the Cayman and Pedro Castle Mem- disconformity that separates the Cayman and Pedro Castle bers are partly or completely filled with lithified and unlithi- Members. Prior to 1984, the south wall of the quarry in- fied terra rossa. The terra rossa was washed into the cavities cluded a vertical section through a cave (22 m long and up from the soils which occur on the surface of the Bluff to 4.5 m high) that was filled with a wide variety of internal 48,50 Fonnation in many parts of the island1. The terra rossa has sediments and speleothems . The floor of that cave lay 9 not been dolomitized. Locally, the cavities are filled with a m below the disconformity that separates the Cayman and breccia formed of angular fragments of white dolostone Pedro Castle Members (Fig. 5.10). During successive years, (derived from the Bluff Formation) held in a terra rossa progressive blasting of the quarry wall meant that the cave matrix. could be traced in the third dimension to the southwest. This Many sinkholes, which are up to 30 m (average 6 m) showed that the original cave was part of an extensive cave deep, are filled or partly filled with breccias. Although some and tunnel system. In 1989, the exposure provided a section of the breccias contains dolostone clasts25 derived from the through a sinkhole that was filled with dolomitized internal Bluff Formation, others contain clasts formed of many sediment (predominantly white and beige caymanite). That different lithologies 37,48. Of particular interest is the fact sinkhole and its fill was truncated by the disconformity (Fig. that many clasts are formed of limestones that have no 5.10). Furthermore, at the disconformity the internal sedi- surface counterparts37. As such they provide a record of ment that filled the sinkhole was bored by the same assem-

96 BRIAN JONES

The Cayman Islands

blage of organisms that bored the upper surface of the thick. surrounding Cayman Member. This clearly indicates that Weathered surfaces of the Ironshore Formation, like the Cayman Member was lithified, karstified and had caves those on the Bluff Formation, are typically black in colour. filled prior to deposition of the Pedro Castle Member. The scalloped nature of these weathered surfaces, as well as Important clues regarding the history of karst on the colour, can be attributed to the diverse array of epilithic and Grand Cayman comes from the internal sediments and endolithic microbes that occur on them27. Tree roots also precipitates that occur in many of the caves, cavities and have a significant effect on the surface exposures of the joints. Examination of hundreds of cavities and caves in the limestones in the Ironshore Formation. The result of tree Bluff Formation on Grand Cayman shows that the cayman-ite roots penetrating the limestone is largely a function of the and grainstones are always dolomitized whereas the terra hardness of that limestone. Roots that penetrate friable, rossa and flowstones are never dolomitized. If all four poorly-consolidated limestones usually have well-devel- cavity-fills are present in a single cave or cavity, the strati- oped rhizoliths around them43. Such rhizoliths developed graphic order is always from caymanite to dolomitized because of the preferential formation of a wide array of grainstone (or interbedded) to flowstone or terra rossa. cements around the roots63. Roots that penetrate hard, well- Although most cavities and caves contain only one or two lithified limestones produced large borings that became the of these fills, no cavity has been found with a sequence that site of intense microbial activity26. Various organisms, such would contradict the above order. This suggests that some as mites, springtails, collembolas and worms are responsible cavities were open and receiving internal sediments both for the formation of peloids in living plant roots that occur before and after dolomitization. The caves that contain only in the limestones of the Ironshore Formation49. flowstones, stalactites, stalagmites and terra rossa probably formed after dolomitization. This suggestion is supported by the fact that the floors of such caves on Cayman Brae can be HOLOCENE SWAMP DEPOSITS traced laterally into the wave-cut notch that formed in association with the sea level highstand approximately Much of Grand Cayman is covered with swamps that 125,000 years ago. 88,90 formed in response to a Holocene transgression . Swamp deposits around Governor's Harbour (Fig. 5.3 A) on the West Peninsula of Grand Cayman, studied in detail by DIAGENETIC FABRICS IN THE IRONSHORE 88,90 92 FORMATION Woodroffe and Woodroffe et al. , are ideal for char - acterizing these deposits (Fig. 5.11). Woodroffe88,89 di- The term ironshore was originally coined because these vided the swamp sediments into mangrove peat, orange limestones characteristically form hard outcrops in coastal plastic mud and green plastic mud. The distribution and areas24,86. In many ways this is a misleading term because the thickness of the black mangrove peat, which is formed of Ironshore Formation is composed mainly of friable and decaying organic matter derived from the plants, is related poorly lithified limestones. The hardness of these lime- to the distribution of the vegetation. The orange plastic mud stones in the coastal settings is a reflection of diagenetic has a low organic content, a relative low moisture content 88 processes that have occurred as a consequence of their (51-85%), and a wet pH of 6.3 to 7.1. Woodroffe sug- exposure to sea water and freshwater. gested that this mud formed on dry ground as a subaerial soil The limestones of the Ironshore Formation contain a beneath a low herbaceous vegetation. The green plastic wide array of fabrics that record diagenesis in the marine, mud, which is non-fibrous with a low organic content and freshwater phreatic and vadose realms (Table 5.3). Vadose relative low moisture content (70-80%), usually occurs in diagenetic fabrics are common, and usually dominate over the lower part of the swamp succession. the marine and freshwater phreatic cements (Table 5.3). In The area to the south of Governor's Harbour (Fig. 5.3 A) many places the limestones of the Ironshore Formation are is covered by a relatively thick (0.5 to 3.0 m) blanket of capped by a well-developed calcrete crust that is up to 6 cm mangrove peat that developed beneath the Rhizophora scrubs of the area88. The thinnest cover occurs near the West

Figure 5.8. (previous page) (A) Localities on Grand Cayman where outcrops of the Ironshore Formation occur (modified after Jones and Hunter36, in which a full list of outcrops appears). (B) Occurrence of facies in the Iron- shore Formation on Grand Cayman (from Hunter and Jones 23). Lack of outcrop in some areas precludes more detailed mapping. (C) Palaeogeographic map of Grand Cayman approximately 125,000 vears ago, when the limestones of the Ironshore Formation were being deposited (from Jones and Hunter34).

98 BRIAN JONES

Facies Lithology Sedimentary Fossils Ichnology Structures Dominant Minor Ichnogenera Burrowing

Bivalve Mudstone to None Bivalves Gastropods None --- wackestone Corals, Algae Foraminifera

Coral A Floatstone None Corals Bivalves Gastrochaenolites --- Gastropods Entobia, Trvpanites

Coral B Floatstone to None Corals Bivalves None --- rudstone Gastropods

Laminated to highly Grainstone Horizontal laminae None Bivalves Skolithos 1 to 10 burrowed grainstone (ooids or Small cross-laminae Ophiomorpha skeletal sand) Conichnus Polvkladichnus Laminated to low Grainstone Horizontal laminae None None None --- angle cross-bedded (ooids or Low angle (<5°) grainstone skeletal sand) cross-bedding

Rudstone Bioclasts or Locally horizontal None Bivalves Qphiomorpha 4 to 10 lithoclasts in laminae or high-angle Corals Polvkladichnus grainstone matrix cross-bedding Conichnus Skolithos

Multidirectional high- Grainstone High angle (25-30°) None Bivalves Conichnus 2 to 3 angle cross-bedded (ooids) cross-bedding Ophiomorpha grainstone Multidirectional

Unidirectional Grainstone High-angle (30°) None None Bergaueria 1 to 3 high-angle cross- (ooids) cross bedding Psilonichnus bedded grainstone Unidirectional

Table 5.1. Characteristic features of facies present in the Ironshore Formation on Grand Cayman (based on Jones and Hunter34; burrowing scale as used in Jones and Pemberton46).

Bay Road where thicknesses of less than 1 m are common. GEOLOGICAL HISTORY OF THE CAYMAN Even in this area, however, ther e are pockets of peat over 2 ISLANDS m thick (Fig. 5.12). To the east, there is an extensive area of A cursory inspection of the geological maps of the Cayman thick (3 m or more) peat cover. Islands showing the Bluff and Ironshore Formations (Fig. Sediment distribution is more complicated to the north 5.3) could easily lead to the conclusion that their geological of Governor's Harbour where mangrove peat occurs with evolution was relatively straightforward and thus easy to orange and green plastic muds (Fig. 5.11). Most mud is interpret. Such a conclusion would be erroneous because the located close to the 'dry cays' that are formed where the simple lithostratigraphic divisions of the strata disguise the Pleistocene bedrock comes to the surface (Fig. 5.11). The true picture. Any consideration of the geological history of mangrove peat of this area is generally less than 1 m thick, these islands must centre around the fact that they have gone although there are isolated pockets up to 2 m thick (Fig. through four phases of deposition, each depositional phase 5.11). The thickness of the peat and muds, which varies being separated by a period of subaerial exposure. In addi- throughout the area, is a function of the amount of relief tion, any model of this type must allow for the dolomitiza- developed on the bedrock prior to the development of the tion of the Bluff Formation and the differential uplift of the swamps (Fig. 5.11). Cores from various parts of the swamp three islands. A major difficulty in any attempt to synthesize show that the swamp sediments are relatively consistent the geological history of these islands lies in the problem of throughout their depth (Fig. 5.12). The orange and green accurately dating the strata, the tuning of the dolomitization plastic muds, which are confined to the lower parts of the and the timing of tectonic activity. For the purposes of this successions, are overlain by mangrove peat (Fig. 5.12). If discussion, the geological evolution of the Cayman Islands the muds are absent, the mangrove peat rests directly on top has been divided into six distinct stages. of the underlying Ironshore Formation (Fig. 5.12). 99 The Cayman Islands

FACIES CORAL A CORAL B LOCATION SB SH DP TF BT CD F MH B K SD I

Stephanocoenia michelini R Acropora palmata S A A S Acropora cervicornis S A A C C Pocillopora sp. C R Agaricia agaricites S S S S C S R S S R Agaricia fragilis R S R R Leptoseris cucullata R Siderastrea siderea S C S C C C S C Siderastrea radians S S Porites astreoides S S S S S S C S S S Porites porites A C C A C A C S C A S A Favia fragum S S S S S S S S S S Diploria clivosa C C S S Diploria strigosa C C C C C C C A C C Diploria lab yrinthiformis S S S C S S C C C S Manicina areolata S S R R S S S R S S S S Col pophyllia natans S S S S S C Montastrea annularis A A A A A A C A A A A C Montastrea cavernosa S C C S C R C Oculina diffusa R R Meandrina meandrites R R R Dichocoenia s tokesi R R R R Dendrogvra c ylindrus C C R S Mussa angulosa R Scol ymia cubensis R R R Isophvllastrea rigida R R R R R Isophvllia sinuosa R R R R R Mycetophvllia spp R S S C S R S Eusmilia fasti giata R R R R R R

Table 5.2. Occurrence of corals at various localities in the Ironshore Formation on Grand Cayman (see Fig. 5.8A). Key: A=abundant; C=common; S=scarce; R=rare.

Stage I by a disconformity that is a record of the subaerial period The oldest strata exposed on the Cayman Islands belong to which followed deposition of the Cayman Member. Dating the Cayman Member of the Bluff Formation (Fig. 5.13). of the Cayman Member and the overlying Pedro Castle Coirelation of this member with other Oligocene strata in Member led Jones and Hunter33 to suggest that this period of the Caribbean suggests that sedimentation terminated in the exposure lasted for approximately 16 Ma. late middle Oligocene33. Indirect evidence suggests that this During this period, limestones of the Cayman Member was about 30 Ma ago33. became fully lithified and extensive leaching of the arago- There is no evidence of reefs in the Cayman Member. nitic fossils occurred. In addition, large-scale dissolution led Indeed, most corals occur as isolated heads or as thickets of to the formation of large caves and passages, the best exam- branching corals. Interpretation of the depositional fabrics ple occurring in the quarry near Pedro Castle. Such cave and faunal components suggests that the Cayman Member systems were subsequently filled with a complex array of originated as carbonates that accumulated on a bank that was sediments (Fig. 5.10). On a smaller scale, cavities were lined probably akin to the Twelve Mile Bank which presently with limpid dolomite crystals, probably as a result of their exists 16 km west of Grand Cayman87. As such, a relatively residence in the mixed-water zone. quiet-water setting with depths of 25 to 40 m is envisaged for most of these deposits. However, in some areas there are Stage III indications of shallower water and higher energy conditions. A transgression in middle Miocene times led to the For example, preliminary work on the Cayman Member of incursion of the sea over the lithified strata of the Cayman Cayman Brae has shown that it includes beds formed of Member. The extent of that transgression and the resultant rhodoids and coral rubble that probably formed in relatively deposits is not known because subsequent weathering shallow water and high energy conditions. Thus, it is possible stripped most of the Pedro Castle Member from Grand that the Cayman Member encompasses a broader spectrum Cayman. Although Jones and Hunter 33 originally thought that of depositional settings than presently suspected. the Pedro Castle Member was restricted to a small area around Pedro Castle, subsequent fieldwork has shown that it covers a Stage II broader area (Fig. 5.3 A). The upper boundary of the Cayman Member is marked The middle Miocene transgression led to marine ero-

100 BRIAN JONES

Figure 5.9. Correlation of sequences through the Ironshore Formation on Grand Cayman (from Jones and Hunter34).

sion of the Cayman Member. Thus, any surface karst fea- Stage IV tures that may have existed on the Cayman Member were The period between deposition of the Pedro Castle eroded to produce a relatively flat, featureless erosional Member and the overlying Ironshore Formation is difficult surface. This erosional surface was extensively bored by to interpret because of the lack of deposits. To date, no late 33 sponges, bivalves and worms . The borings are a clear Miocene, Pliocene, or early to middle Pleistocene strata indication that the substrate was fully lithified prior to have been found on the Cayman Islands. Thus, there is a deposition of the Pedro Castle Member. This erosional period of approximately 15 Ma which poses a problem for surface also cut through the limestones that filled the sink- any discussion concerning the geological evolution of the holes which had previously developed in the Cayman Mem- Cayman Islands. ber. Indeed, the erosional surface which cuts the It is evident, however, that this period saw (i) the sinkhole-filling deposits at Pedro Castle Quarry is also dolomitization of the limestones in the Bluff Formation; (ii) marked by borings that are identical to those penetrating the differential uplift of the three Cayman Islands; and (iii) surrounding bedrock (Fig. 5.10). weathering of the Bluff Formation. The basal bed of the Pedro Castle Member is a grain- Dolomitization stone that commonly contains numerous rhodoids. The beds Consideration of available petrographic and geochemi- overlying the grainstone are characterized by numerous cal data suggests that there was only one phase of dolomiti- free-living corals and only rare colonial corals. The basal zation 2 to 5 ma ago33,39,66,68. The process responsible for grainstones and rhodoids probably formed in a relatively dolomitization of the Bluff Formation is not known with shallow, high-energy setting. This interpretation is reason- certainty. Jones et al.39 suggested that it occurred in the able because they are the initial record of a sea transgressing mixed-water zone; the pervasive nature of the dolomitiza- over a hard substrate. The environmental significance of the tion being due to fluctuating sea levels and the related abundant, large free-living corals is not fully understood. changes in the position of the mixed-water zone. Ng62 Weathering over the last 15 Ma has removed much of agreed with this suggestion, but argued that there were two the Pedro Castle Member. It is thus difficult to establish the phases of dolomitization, one following deposition of the thickness of the member or to determine the time when this Cayman Member and the second after deposition of the phase of deposition terminated. Pedro Castle Member. Conversely, Pleydell et al6S argued

101 The Cayman Islands

Figure 5.10. Schematic diagrams showing cave that was present in the southwest wall of the quarry near Pedro Castle (see Fig. 5.3 A). Inset map shows the position of the quarry wall in 1983, 1989 and 1990. (A) Cross-section in 1983 showing through a cave that was filled with caymanite, dolomitized grainstone, flowstone and terra rosa (modified after Lockhart j. (B) Cross-section in 1989 after wall of quarry had been excavated about 20 m southwest from the 1983 position. The cave shown in this diagram is the southwest extension of the cave shown in (A).

that natural sea water was probably responsible for dolomi- gence would have occurred 2 to 5 ma ago. There is, however, tization of the Bluff Formation. no physical evidence of such submergence because there are If seawater is deemed responsible for dolomitization of no strata of that age. This may reflect that (i) this was a the Bluff Formation, then there must have been a period period of submergence, but non-deposition; (ii) such strata when each of the Cayman Islands were submerged. Inter - were deposited and subsequently removed by erosion; or pretation of the Sr isotope data indicates that this submer- (iii) such strata are present but have not yet been recognized.

102 BRIAN JONES

Figure 5.11. Distribution of swamp deposits in the area north of Governor's Harbour (West Peninsula, Grand Cayman). Note lateral variation in the thickness of the swamp deposits that are related to the topography developed on the underlying Ironshore Formation (compiled and modified from information in Woodroffe88).

If the idea of dolomitization by seawater is rejected then this deposited. Uplift of Cayman Brae must have occurred after process must be related to the mixed-water zone because the dolomitization of the Bluff Formation, but prior to the late geological setting of these islands precludes the use of any Pleistocene highstand 125,000 years ago. other dolomitization model. Weathering of the Bluff Formation Differential uplift of the Cayman Islands During the period between the deposition of the Pedro Grand Cayman and Little Cayman are similar in terms Castle Member and the Ironshore Formation there was of their surface topography and elevation above sea level. extensive subaerial weathering of the Cayman Islands. As - However, Cayman Brae is notably different because it rises sessment of this weathering is hindered because it is difficult to nearly 50 m at its northeastern end. The strata on that to segregate its effects from more recent weathering. Nev- island appear to dip to the southwest. Dating the uplift of the ertheless, it is evident that this phase of erosion led to the Cayman Brae relative to the other islands is constrained by removal of the Pedro Castle Member from large areas of two considerations. First, uplift must have postdated Grand Cayman and the formation of a rugged karst terrain dolomitization because even the highest strata on Cayman on the rocks of the Bluff Formation. Brae are dolomitized, and yield the same petrographic and The topography developed on the Bluff Formation prior geochemical signatures as the dolostones of this formation to deposition of the Ironshore Formation is evident from the on Grand Cayman and Little Cayman. Secondly, on Cayman relief that presently exists on the unconformity that sepa- Brae there is a distinct wave-cut notch at 6 m above sea level rates the two formations. In some areas the unconformity is which is cut into the dipping strata of the Bluff Formation. essentially horizontal (for example, Newlands and south of This wave-cut notch, which correlates with a similar wave- George Town) whereas in other areas it is vertical or sub- cut notch on Grand Cayman, formed 125,000 years ago vertical (such as Spots). Locally, the relief on the unconfor- when the limestones of the Ironshore Formation were being mity appears to be as much as 30 m34. It is difficult,

103 The Cayman Islands

Cement Type Abundance Environment

Isopachous aragonite Common in cavities in fossils, incommon some grainstones Marine Peloids Common in cavities in corals Marine Blocky spar calcite Common in some grainstones Freshwater Meniscus cement Common in grainstones Vadosephreatic Needle-fibre cement Common in cavities in fossils; in root borings Vadose Microstalactites Rare Vadose Grain coating needle Common in some grainstones Vadose Calcifiedmats algae/fungi Common in root borings; common in some grainstones Vadose Calcified root hairs Common Vadose Calcified bacteria Common Mostly vadose Pyrite framboids Very rare Vadose?

Table 5.3. Summary of diagenetic fabrics present in the Ironshore Formation on Grand Cayman.

however, to accurately map the topography on the uncon- which has accentuated the karst topography on the Bluff and formity because it is largely buried and below present-day Ironshore Formations. Although the amount of rock re- sea level. moved by weathering is difficult to evaluate, it is probably substantial. For example, in most areas of Grand Cayman Stage V and Little Cayman, weathering has removed the old cliff This stage, which occurred about 125,000 years ago, lines and the wave-cut notches that formed during the last involved deposition of the limestones (Ironshore Forma- highstand. tion) when sea level rose to approximately 6 m above that Approximately 50% of Grand Cayman is presently of the present-day34. The palaeogeography at that time was covered by mangrove swamps90. The sediments beneath controlled by the topography previously developed on each these swamps, which includes peat layers up to 6 m thick, of the islands and the position of sea level. are the record of Holocene submergence of the island90. On Grand Cayman, deposition occurred in a large lar Radiocarbon dates from the lower parts of the peat layers goon (Ironshore Lagoon23) that covered much of the west- suggest that this transgression was initiated approximately ern half of the island (Fig. 5.8B), and in small lagoons that 2100 years ago89. occurred around the south, east and north coasts of the island Present-day reefs and carbonate sedimentation occur in (Fig. 5.8B). Analysis of the limestones in the Ironshore the shallow-water sounds and on narrow shelves around Formation showed that the Ironshore Lagoon incorporated each of the Cayman Islands. Various aspects of the modern a bivalve facies, a belt of patch reefs and a barrier reef sedimentation patterns have been documented18,32,59,71,72, across its western entrance (Fig. 5.8B). 73,77,80. Beachrock, which is common at many localities On Cayman Brae, limestones of the Ironshore Forma- around the coast of Grand Cayman, has been described by 57,58 30 tion were deposited in narrow lagoons and shelves that Moore and Jones and Goodbody . fringed the upstanding block of the Bluff Formation. Asso- ciated with the well-developed wave-cut notch which occurs ACKNOWLEDGEMENTS—-The research on which this paper is based would have been impossible without the financial support provided by the around the island are numerous caves. Natural Science and Engineering Research Council of Canada (grant No. The extent to which the Sangamon (125,000 ma ago) A67090), and the logistical assistance given by Mr. Richard Beswick, highstand submerged Little Cayman is difficult to evaluate Director, Water Authority, Cayman Islands and Dr. J. Davies, Director, Mosquito Research Control Unit, Cayman Islands. I am indebted to these because subsequent erosion has removed much of the se- organizations for their support. I am grateful to my graduate students, quence. Nevertheless, it would appear that the situation on Cheryl Squair, Blair Lockhart, Suzanne Pleydell, Duncan Smith, Jill Little Cayman largely paralleled that on Grand Cayman. Rehman, Sam Ng and Ian Hunter, who have all allowed me to use information gathered during their studies of the Cayman Islands. I am also Stage VI indebted to these students for their help in the field. SamNg and Ian Hunter critically reviewed, and Elsie Tsang helped me prepare, this manuscript. Over the last 125,000 years, the topography of each of Two anonymous reviewers also posed many questions that forced clarifi- the Cayman Islands has been modified by weathering75 cation of certain points.

104

BRIAN JONES

Figure 5.12. Sediment sequences in swamp deposits in the area north of Governor's Harbour (modified after Woodroffe88). & Richards, H.G. 1973. The Pleistocene rocks of the REFERENCES Cayman Islands. Geological Magazine, 110, 209-221. 6 1 Bugg, S.F. & Lloyd, J. W. 1976. A study of freshwater lens Ahmad, N. & Jones, R.L. 1969. Occurrence of aluminous configuration in the Cayman Islands using resistivity lateritic soils (bauxites) in the Bahamas and Cayman methods. Quarterly Journal of Engineering Geology, 9, Islands. Economic Geology, 64, 804-808. 2 291-302. Beckmann, J.P., Butterlin, J., Cederstnom, D.J., Christman, 7 Carew, J.L. 1984. Estimates of late Pleistocene sea level high R.A., Chubb, L.J., Hoffstetter, R., Kugler, H.G., Mar- tin-Kaye, P.H.A., Maxwell, J.C., Mitchell, R.C., stands from San Salvador, Bahamas: in Mylroie, J. (ed.), Ramirez, R., Versey, H.R., Weaver, J.D., Westermann, Proceedings of the Second Symposium on the Geology of J.H. &Zans, V.A. 1956. LexiqueStratigraphique Inter- the Bahamas, 17th~22nd June, 153-177. Bahamian Field national, Volume 5. Amerique Latine, 2b, Antilles. Station, Fort Lauderdale. 8Cenidwen, S.A. 1989. Paleoecology of Pleistocene Mol- CNRS, Paris, 495 pp. 3Bowin, C.O. 1968. Geophysical study of the Cayman luscafrom the Ironshore Formation, Grand Cayman, B. Trough. Journal of Geophysical Research, 73, 5159- W.I. Unpublished M.Sc. thesis, University of Alberta, 271pp. 5173. 9 4Brasier, M. & Donahue, J. 1985. Barbuda—an Cerridwen, S.A. & Jones, B. 1991. Distribution of bivalves and emerging reef and lagoon complex on the edge of gastropods in the Pleistocene Ironshore Formation, Grand the Lesser Antilles island arc. Journal of the Cayman, British West Indies. Caribbean Journal of Science, 27, 97-116. Geological Society, London, 142, 1101-1117. 10 5 Dillon, W.P., Vedder, J.G. & Graf, R.J. 1972. Structural Brunt, M.A., Giglioli, M.E.C., Mather, J.D., Piper, D.J.W. 105 The Cayman Islands

JAMAICA ANGUILLA- GRAND NORTH/ CLARENDON ANTIGUA CAYMAN CENTRAL BLOCK BANK

Figure 5.13. Stratigraphic columns showing correlation of the Bluff Formation with other sequences of similar age in the Caribbean region (modified after Jones and Hunter 33).

profile of the northwestern Caribbean. Earth and 13Fahlquist, D.A. & Davies, D.K. 1971. Fault-block origin Planetary Science Letters, 17, 175-180. of western Cayman Ridge, Caribbean Sea. Deep-Sea 11Doran, E. 1954. Landforms of Grand Cayman Island, Research, 18, 243-253. British West Indies. Texas Journal of Science, 6,360- 14Folk, R.L. & McBride, E.F. 1976. The Caballos Novacu- 377. lite revisited—part 1: origin of novaculite members. 12Emery, K.O. & Milliman, J.D. 1980. ShaUow-water lime- Journal of Sedimentary Petrology, 46, 659-669. stones from slope off Grand Cayman Island Journal of 15Folk, R.L., Roberts, H.H. & Moore, C.H. 1973. Black Geology, 88, 483-488. phytokarst from Hell, Cayman Islands, British West

106 BRIAN JONES

Indies. Geological Society of America Bulletin, 84, Earth Sciences, 28, 382-397. 2351-2360. 29Jones, B. 1992. Caymanite—a cavity filling deposit in the 16Frost, S.H. & Langenheim, R.L. 1974. Cenozoic reef Oligocene-Miocene Bluff Formation of the Cayman biofacies. Tertiary larger Foraminifera and scleract- Islands. Canadian Journal of Earth Sciences. 29, 720- inian corals from Chiapas, Mexico. Northern Illinois 736. University Press, Dekalb, 388 pp. 30Jones, B. & Goodbody, Q.H. 1982. The geological signifi- 17Harmon, R.S., Land, L.S., Mitterer, R.M., Garrett, P., cance of endolithic algae in glass. Canadian Journal of Schwarcz,H.P, & Larson, GJ. 1981. Bermuda sea level Earth Sciences, 19, 671-678. during the last interglacial. Nature, 289, 481-483. 31Jones, B. & Goodbody, Q.H. 1984. Biological factors in 18Hernandez-Avila, M.L. & Roberts, H.H. 1977. Hurricane- the formation of quiet water ooids. Bulletin of Cana- generated waves and coastal boulder rampart forma- dian Petroleum Geology, 32, 190-199. tion: in Taylor, D.L. (ed.). Proceedings of the Third 32Jones, B. & Goodbody, Q.H. 1985. Oncolites from a International Coral Reef Symposium, University of Mi- shallow water lagoon, Grand Cayman Island. Bulletin ami, Miami, May, 71-78. University of Miami, Miami. of Canadian Petroleum Geology, 32, 254-260. 19Holcombe, P.R.V., Matthews, J.E. & Murchison, R.R. 33Jones, B. & Hunter, I.G. 1989. The Oligocene-Miocene 1973. Evidence for sea-floor spreading in the Cayman Bluff Formation on Grand Cayman. Caribbean Journal Trough. Earth and Planetary Science Letters, 20, 357- of Science, 25, 71-85. 34 371. Jones, B. & Hunter, I.G. 1990. Pleistoc ene paleogeogra- 20Horsfield, W.T. 1975. Quaternary vertical movements in phy and sea levels on the Cayman Islands, British West the Greater Antilles. Geological Society of America Indies. Coral Reefs, 9, 81-91. Bulletin, 86, 933-938. 35Jones, B. & Hunter, I.G. 1991. Corals to rhodolites to 21Hunter,I.G. & Jones, B. 1988. Corals and paleogeography microbialites —a community replacement sequence in- of the Pleistocene Ironshore Formation on Grand Cay- dicative of regressive conditions. Palaios, 6, 54-66. man, B.W.I.: in Choat, J.H. et al. (eds.), Proceedings of 36Jones, B., Hunter, I.G. & Ng, K.C. 1990. Geological the Sixth International Coral Reef Symposium, Towns- evolution of the Oligocene-Miocene Bluff Formation, ville, Australia, 8th-J2th August, 3,431-435. 6th ICRS Grand Cayman: in Lame, D.K. & Draper, G. (eds), Executive Committee, Townsville. Transactions of the Twelth Caribbean Geological Con- 22Hunter. I.G. & Jones, B. 1989. Sedimentology of the late ference, St. Croix, U.S. Virgin Islands, 7th-llth August, Pleistocene Ironshore Formation on Grand Cayman. 1989, 125-132. Abstracts, Twelfth Caribbean Geological Conference 37Jones, B. & Kahle, C.F. 1985. Lichen and algae: agents of St Croix, U.S. Virgin Islands, 7th-llth August, p. 78. biodiagenesis in karst breccia from Grand Cayman 23Hunter, I.G. & Jones, B. 1990. Sedimentology of the late Island. Bulletin of Canadian Petroleum Geology, 33, Pleistocene Ironshore Formation on Grand Cayman: in 446-461. Lame, D.K. & Draper, G. (eds), Transactions of the 38Jones, B. & Kahle, C.F. 1986. Dendritic calcite crystals Twelth Caribbean Geological Conference, St. Croix, formed by calcification of algal filaments in a vadose U.S. Virgin Islands, 7th-11th August, 1989,104-114. environment. Journal of Sedimentary Petrology, 56, 24 Idyll, C.P. 1966. The Ironshore. Sea Frontiers, 12, 328- 217-227. 339. 39Jones, B., Lockhart, E.B. & Squair, C. 1984. Phreatic and 25Jones, B. 1987. The alteration of sparry calcite crystals in vadose cements in the Tertiary Bluff Formation of a vadose setting, Grand Cayman Island. Canadian Grand Cayman Island, British West Indies. Bulletin of Journal of Earth Sciences, 24, 2292-2304. Canadian Petroleum Geology, 32, 382-397. 26Jones, B. 1988. The influence of plants and micro-organ- 40Jones, B. & MacDonald, R.W. 1989. Micro-organisms isms on diagenesis in caliche: example from the and crystal fabrics in cave pisoliths from Grand Cay- Pleistocene Ironshore Formation on Cayman Brae, man, British West Indies. Journal of Sedimentary Pe- British West Indies. Bulletin of Canadian Petroleum trology, 59, 387-396. Geology, 36,191-201. 41Jones, B. & Motyka, A. 1987. Biogenic structures and 27 Jones, B. 1989. The role of micro-organisms in phytokarst micrite in stalactites from Grand Cayman Island, Brit- development on dolostones and limestones, Grand ish West Indies. Canadian Journal of Earth Sciences, Cayman, British West Indies. Canadian Journal of 24,1402-1411. Earth Sciences, 26, 2204-2213. 42Jones, B. & Ng, K.C. 1988a. Anatomy and diagensis of a 28 Jones, B. 1991. Genesis of terrestrial oncoids, Cayman Pleistocene carbonate breccia formed by the collapse of Islands, British West Indies. Canadian Journal of a seacliff, Cayman Brae, British West Indies. Bulletin

107 The Cayman Islands

of Canadian Petroleum Geology, 36,9-24. (British West Indies) and their relation to the Bartlett 43Jones, B. & Ng, K.C. 1988b. The structure and diagenesis Trough. Quarterly Journal of the Geological Society, of rhizoliths from Cayman Brae, British West Indies. London, 82, 352-387. 56 Journal of Sedimentary Petrology, 58, 457-467. Matthews, R.K. 1973. Relative elevation of late Pleisto- 44Jones, B. & Pemberton, S.G. 1988a. Bioerosion of corals cene high sea level stands. Quaternary Research, 3, byLithophaga: example from the Pleistocene Ironshore 147-153. Formation of Grand Cayman, B.W.I. Proceedings of 57Moore, C.H. 1971. Beach rock cements, Grand Cayman the Sixth International Coral Reef Symposium, Towns- Island, B.W.I.: in Bricker, O.E. (ed.), Carbonate Ce- ville, Australia, 8th-12th August, 3,437-440.6th ICRS ments. Johns Hopkins University Press, Baltimore, 9- Executive Committee, Townsville. 12. 45Jones, B. & Pemberton, S.G. 1988b. Lithophaga borings 58Moore, C.H. 1973. Intertidal carbonate cementation, and their influence on the diagenesis of corals in the Grand Cayman, West Indies. Journal of Sedimentary Pleistocene Ironshore Formation of Grand Cayman Is- Petrology, 43, 591-602. land, British West Indies. Palaios, 3, 3-21. 59Murray, S.P., Roberts, H.H., Conlon, D.M. & Rudder, 46Jones, B. & Pemberton, S.G. 1989. Sedimentology and G.M. 1977. Nearshore current fields around coral is- ichnology of a Pleistocene unconformity-bounded, lands: control on sediment accumulation and reef shallowing upward carbonate sequence: the Ironshore growth: in Taylor, D.L. (ed.), Proceedings of the Third Formation, Salt Creek, Grand Cayman. Palaios, 4, International Coral Reef Symposium, University of Mi- 343-355. ami, Miami, May, 54-59. University of Miami, Miami. 47Jones, B, Pleydell, S.A, Ng, K.C. & Longstaffe, F.J. 60Neumann, A. & Moore, W.S. 1975. Sea level events and 1989. Formation of poikilotopic calcite-dolomite fab- Pleistocene ages in the northern Bahamas. Quaternary rics in the Oligocene-Miocene Bluff Formation of Research, 5, 215-224. Grand Cayman, British West Indies. Bulletin of Cana- 61Ng, K.C. 1985. Geological aspects of ground water ex- dian Petroleum Geology, 37, 255-265. ploitation in Grand Cayman. United Nations Interre- 48Jones, B. & Smith, D.S. 1988. Open and filled karst gional Seminar on Development and Management of features on the Cayman Islands: implications for the Island Groundwater Resources, 2nd-6th December, recognition of paleokarst. Canadian Journal of Earth 6.1-6.29. Government of Bermuda, Hamilton. Sciences, 25, 1277-1291. 62Ng, K.C. 1990. Diagenesis of the Oligocene-Miocene 49Jones, B. & Squair, C.A. 1989. Formation of peloids in Bluff Formation of the Cayman Islands—a pet- plant rootlets, Grand Cayman, British West Indies. rographic and hydrogeochemical approach. Unpub- Journal of Sedimentary Petrology, 59, 1002-1007. lished Ph.D. thesis, University of Alberta, 344 pp. 50Lockhart, E.B. 1986. Nature and genesis ofcaymanite in 63Ng, K.C. & Jones, B. 1990. Porosity and permeability the Oligocene-Miocene Bluff Formation of Grand development in the Oligocene-Miocene BluffForma- Cayman, British West Indies. Unpublished M.Sc. tion, Grand Cayman: in Larue, D.K. & Draper, G. (eds), thesis, University of Alberta, 111 pp. Transactions of the Twelth Caribbean Geological Con- 51 MacDonald, K.C. & Holcombe, T.L. 1978. Inversion of ference, St. Croix, U.S. Virgin Islands, 7th- llth August, magnetic anomalies and sea-floor spreading in the Cay- 1990,115-124. man Trough. Earth and Planetary Science Letters, 40, 64Pemberton, S.G. & Jones, B. 1988. Ichnology of the 407-414. Pleistocene Ironshore Formation, Grand Cayman Is- 52 Malin, P.E. & Dillon, W.P. 1973. Geophysical reconnais- land, British West Indies. Journal of Paleontology, 62, sance of the western Cayman Ridge. Journal of Geo- 495-505. physical Research, 78, 7769-7775. 65Perfit, M.R. & Heezen, B.C. 1978. The geology and 53 Mather, J.D. 1972. The geology of Grand Cayman and its evolution of the Cayman Trench. Geological Society of control over the development of lenses of potable America Bulletin, 89, 1155-1174. groundwater. in Petzall, C. (ed.), Transactions of the 66Pleydell, S.M. 1987. Aspects of diagenesis and ichnology in Sixth Caribbean Geological Conference, Margarita the Oligocene-Miocene BluffFormation of Grand Island, Venezuela, 6th-14thJuly, 1971,154-157. Cayman Island, British West Indies. Unpublished 54 Matley, C.A. 1925. A reconnaissance geological survey M.Sc. thesis, University of Alberta, 209 pp. of the Cayman Islands, B.W.I. Report of the British 67Pleydell, S.M. & Jones, B. 1988. Boring of various faunal Association for the Advancement of Science, Toronto, elements in the Oligocene-Miocene BluffFormation 1924, 392-393. of Grand Cayman, British West Indies. Journal of 55Matley, C.A. 1926. The geology of the Cayman Islands Paleontology, 62, 348-367.

108 BRIAN JONES

68Pleydell, S.M., Jones, B., Longstaffe, FJ. & Baadsgaard, pentaria from the Cayman Islands and their H. 1990. Dolomitization of the Oligocene-Miocene geological significance. Geological Magazine, 82, Bluff Formation on Grand Cayman, British West In- 388-400. dies. Canadian Journal of Earth Sciences, 27, 1098- 83Vilas, H.A. & Spencer, T. 1983. Phytokarst, blue-green 1110. algae and limestone weathering: in Paterson, K. & 69Rehder, H.A. 1962. The Pleistocene mollusks of Grand Sweeting, M.M. (eds), New Directions in Karst: Cayman Island, with notes on the geology of the island. Proceedings of the Anglo-French Karst Journal of Paleontology, 36, 583-585. Symposium, September 1983, 115-140. 70Richards, H.G. 1955. The geological history of the Cay- 84Ward, W.C. 1985. Quaternary geology of northwestern man Islands. Notulae Naturae of the Academy of Natu- Yucatan Peninsula: in Ward, W.C., Weidie, A.E. & ral Sciences of Philadelphia, 284, 1-11. Back, W. (eds), Geology and Hydrogeology of the 71Rigby, J.K. & Roberts, H.H. 1976. Geology, reefs and Yucatan and Quaternary Geology ofNortheastem marine communities of Grand Cayman Island, B.W.I. Yucatan Peninsula, 23-95. New Orleans Geological Geology Studies of Brigham Young University, Special Society, New Orleans. Publication, 4,122 pp. 85Ward, W.C. & Brady, M.J. 1979. Strandline sedimenta- 72Roberts, H.H. 1971. Environments and organic commu- tion of carbonate grainstones, Upper Pleistocene, nities of North Sound, Grand Cayman, B.W.I. Carib- Yu catan Peninsula, Mexico. American Association bean Journal of Science, 2, 67-79. of Petroleum Geologists Bulletin, 63,362-369. 73Roberts, H.H. 1977. Field Guidebook to the Reefs and 86Warthin, A.S. 1959. Ironshore in some West Indian is- Geology of Grand Cayman Island, B. W.L Third Inter- lands. Transactions of the New York Academy of national Coral Reef Symposium, University of Miami, Science, 21, 649-652. Miami, 41 pp. 87Wells, S.M. 1988. Coral Reefs of the World. Volume 74Rosencrantz, E., Ross, M.I. & Sclater, J.G. 1988. Age and 1: Atlantic and eastern Pacific. United Nations spreading history of the Cayman Trough as determined Environmental Programme (UNEP), International from depth, heat flow, and magnetic anomalies. Journal Union for Conservation of Nature and Natural of Geophysical Research, 93, 2141-2157. Resources (IUCN), 75Spencer, T. 1985. Weathering rates on a Caribbean reef 88Woodroffe, C.D. 1979. Mangrove swamp stratigraphy limestone: results and implications. Marine Geology, and Holocene transgression, Grand Cayman Island, 69, 195-201. West Indies. Unpublished Ph.D. thesis, University 76Stoddart, D.R. 1980. Geology and geomorphology of of Cambridge, 611 pp. Little Cayman. Atoll Research Bulletin, 241, 11-16. 89Woodroffe, C.D. 1981. Mangrove swamp stratigraphy 77Suhayda, J.N. & Roberts, H.H. 1977. Wave action and and Holocene transgression, Grand Cayman Island, sediment transport on fringing reefs: in Taylor, D.L. West Indies. Marine Geology, 41, 271-294. (ed), Proceedings of the Third International Coral Reef 90Woodroffe, C.D. 1982. Geomorphology and development Symposium, University of Miami, Miami, May, 65-70. of mangrove swamps, Grand Cayman Island, West University of Miami, Miami. Indies. Bulletin of Marine Science, 32, 381-398. 91 78Sykes, L.R., McCann, W.R. & Kafka, A.L. 1982. Motion Woodroffe, C.D. 1988. Vertical movements of isolated of Caribbean Plate during the last 7 million years and oceanic islands at plate margins. Zeitschrift fir Geo- implications for earlier Cenozoic movements. Journal morphologie, 69, 17-37. of Geophysical Research, 87, 10656-10676. 92Woodroffe, C.D., Stoddart, D.R. & Giglioli, M.E.C. 1980. 79Taber, S. 1922. The Great Fault Troughs of the Antilles. Pleistocene patch reefs and Holocene swamp morphol- Journal of Geology, 30, 89-114. ogy, Grand Cayman, West Indies. Journal of Bio- 80Tongpenyai, B. & Jones, B. 1991. Application of image geography, 7, 103-113. analysis for delineating modern carbonate facies 93Woodroffe, C.D., Stoddart, D.R., Harmon, R.S. & changes through time: Grand Cayman, western Carib- Spencer, T. 1983. Coastal morphology and late Quater- bean Sea. Marine Geology, 96, 85-101. nary history, Cayman Islands, West Indies. Quaternary 81Uchupi, E. 1975. Physiography of the Gulf of Mexico and Research, 19, 64-84. Caribbean Sea: in Nairn, A.E.M. & Stehli, F.G. (eds), 94Zans, V.A, Chubb, L.J, Versey, H.R., Williams, J.B., The Ocean Basins and Margins. 3. The Gulf of Mexico Robinson, E. & Cooke, D.L. 1963. Synopsis of the and the Caribbean. Plenum, New York, 1-64. geology of Jamaica. Bulletin of the Geological Survey 82Vaughan, T.W. 1926. Species of Lepidocyclina and Car- Department, Jamaica, 4 (for 1962), 72 pp.

109

CHAPTER 6

Jamaica

EDWARD ROBINSON

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION tive. The earlier work of the present Geological Survey was reported by Zans et al117 . An extensive bibliography JAMAICA, THE third largest of the Greater Antillean is- through 1974 was published by Kinghorn63. Geological lands, with an area of 11,264 square km, plus some 9,600 mapping on a scale of 1:50,000 covers most of the island square km of offshore banks and shoals, lies at the eastern and a 1:250,000 scale map of the whole island has been end of the Nicaraguan Rise, a topographic extension of published74. A general text on the island's rocks, minerals northern Central America. To the north it is bordered by the and geological history is available81, as is a commentary on Cayman Trough, at 7,200 m the deepest spot in the Carib- geological features of local interest, written for the non-spe- bean. To the south is the Colombian Basin (Fig. 6.1). cialist80. Other recent geological summaries include those Some aspects of Jamaica's geology are like those of the by Meyerhoff and Krieg76, emphasizing petroleum pros- other Greater Antillean islands though, as the easternmost pects; Draper31, covering tectonic evolution; and a volume point on the Nic araguan Rise, it also possesses some geo- on the biostratigraphy of Jamaica115. Papers and transact- logical characteristics in common with that region. Jamaica tions of professional meetings on Jamaica's geology are may be regarded as the emergent, uplifted, easterly tip of the published regularly by the Geological Society of Jamaica Nic araguan Rise, dissected to display geological features (such as Ahmad1). Data from the CIDA/GOJ Minerals which elsewhere on the Rise are submerged beneath the sea Project includes several geophysical, structural and geo- and buried beneath a cover of Quaternary carbonate sedi- chemical maps of the island20. ments and rocks (Fig. 6.1). The island's interior is mountainous. In the east a string of peaks, the Blue Mountains, rises up to as much as 2,255 MORPHOTECTONIC FEATURES m in altitude. Over the rest of the island, an extensively dissected limestone plateau reaches a height of over 900 m The morphotectonic features of Jamaica are defined today and displays a rugged karst scenery, the most spectacular by two main sets of onshore faults. An east-west set, on being the famed Cockpit Country101. Along its northern which sinistral strike-slip offsets have been measured or flank this plateau descends abruptly to the coast, in a series inferred, intersects a northnorthwest-southsoutheast trend- of steps controlled by east-west faults and by raised marine ing set, particularly evident over the karst uplands of central terraces. These steps continue steeply offshore down into Jamaica49,107. From east to west, this second set of faults the Cayman Trough. The southerly slopes of the plateau are defines three positive morphotectonic units, the Blue Moun- less steep and are bounded by more or less extensive alluvial tains, Clarendon and Hanover Blocks, which consist of plains, while south of the central part of the island a shallow Cretaceous rocks, mainly of volcanic derivation, covered by marine platform extends up to a maximum of 30 km off- relatively thin, undeformed Tertiary limestones. They are shore. separated by the Wagwater and Montpelier-Newmarket Following the first geological investigation of Jamaica Belts that contain thicker sequences of deformed Tertiary by de la Beche7, many geologists have examined various sedimentary rocks. A third belt, the John Crow Mountains aspects of Jamaican geology and the literature is exhaus - Belt, borders the eastern side of the Blue Mountains Block,

111

Jamaica

Figure 6.1. Simplified geological map of Jamaica (modified from Donovan26). along the east coast of Jamaica. ments indicative of a subduction complex30. These units are truncated to the north by the North Coast Offshore to the south, the block and belt structure, Belt, where the east-west set of faults is dominant (Duanvale defined by the Tertiary onshore fault sets, is at least partly Fault System), and along the south coast of the island by the replaced by a fracture pattern in which northeasterly trends South Coast Fault System (Fig. 6.2). In the extreme south- are dominant51,91. The trends are similar to those perceived west of the island, the Negril - Savanna-la-Mar Belt has also on a regional scale for the Nicaraguan Rise16. Although the been recognised76. In eastern Jamaica the Wagwaler Belt is boundary between the two sets of trends is marked by the continued into the southeast corner of the island, south of South Coast Fault System, the orientation of magnetic the Yallahs-Plantain Garden Fault System. Eva and anomalies over central Jamaica30 indicates a northeastward McFarlane34 reviewed the geological setting of the blocks continuation of the Nic araguan Rise trend into the onshore and belts in Jamaica. Cretaceous, also evident from SEASAT radar imagery107. These structural units, defined by faulting which partly controlled and partly post-dated deposition of most of the Tertiary rocks, are superimposed on a Cretaceous substrate STRATIGRAPHY that can be examined in outcrop within twenty seven inliers, exposed through erosion of the Tertiary cover. The substrate is The stratigraphy of Jamaica is simple in outline, but com- divisible into a western part, dominated by volcaniclastic plex in detail. The Cretaceous succession is dominated by rocks with minor limestones; a central, volcanic region with volcanogenic rocks, locally including flows, with lime- substantial lava flows, pyroclastic deposits and plutons; and stones playing a subordinate role in the Aptian-Albian and an eastern region containing metamorphic and ophiolitic the late Campanian and Maastrichtian. The early Cretaceous slices within a thick, volcanogenic sequence. These three carbonates are associated with volcanic piles of submarine regions have been interpreted as representing, respectively, origin. The late Cretaceous carbonates mainly represent a back-arc basin22,33, a central, volcanic arc, and an eastern, transitional facies in the infilling of the back-arc basin to the fore-arc basin which has been complexly faulted, with ele- west.

112 EDWARD ROBINSON

The Tertiary succession is marked by basal terrigenous rallly isolated metamorphic rocks are juxtaposed against elastics, representing erosion of an apparently widespread Upper Cretaceous volcanic and sedimentary sequences ex- Maastrichtian-Paleocene land area in the central and west- hibiting facies contrasts between ophiolitic rocks in the ern parts of the island. In the east these are associated with southeast of the inlier, and calc -alkaline volcanics, intruded thick, coarse-grained clastic accumulation, keratophyric by granodiorite, over the rest of the inlier110. flows and syntectonic faulting within a possible interarc The small extension of the inlier south of the Plantain basin55. Garden Fault (Sunning Hill district; SU on Fig. 6.3) contains There is transition upsection to an island-wide lime- Upper Cretaceous rocks resembling the calc -alkaline vol- stone sequence. In central and western Jamaica, and con- canic sequence of the main part of the inlier, rather than the tinuing into the western part of the Nicaraguan Rise90, the ophiolitic complex, but the total thickness is much less109. transition is seen in the Middle Eocene Yellow Limestone About 2 to 3 km of downthrow are required on the Plantain Group (Fig. 6.1). In the John Crow Mountains Belt this Garden Fault System for the Cretaceous structural offsets. transition takes place in Paleocene rocks. Carbonate rocks The metamorphic rocks along the southern margin of (White Limestone Supergroup; Fig. 6.1) make up the bulk the Blue Mountains consist of mafic blueschists, green- of the Tertiary section, but are progressively replaced, from schists and rocks of amphibolite facies29,32,62 (Mt. Hibernia the Middle Miocene, by the mixed carbonate and clastic and Westphalia schists; Fig. 6.4 herein). These have gener - Coastal Group (Fig. 6.1), outcropping on the margins of the ally been considered as the oldest rocks in the inlier and by island. Minor pillowed volcanic flows are seen at one locality some76 as the oldest rocks in the island. Lewis et al.76 (Low Layton; Fig. 6.3) in rocks of Miocene age106 and reported isotopic (K/Ar) ages of 76.5 ±2.1 Ma on horn- volcanics of possibly similar age have been dredged off the blende and 48.8 ± 1.3 and 52.9 ± 1.4 Ma on mica from the southeastern tip of the island96 (Fig. 6.3). Over the central amphibolite facies rocks, which Draper29,31 interpreted as uplands the White Limestone is succeeded by superficial being consistent with the thermal evolution of rocks which deposits of terra rossa, in most places sufficiently rich in crystallized originally in an early Cretaceous subduction aluminium minerals to be mined as bauxite46,116. complex. The oldest, palaeontologically-dated rocks in the Blue CRETACEOUS Mountains Inlier are Upper Cretaceous and form part of the Back Rio Grande Group64 (Fig. 6.4). In this unit volcano- Knowledge of the Cretaceous geology of Jamaica is derived genic sandstone and conglomerate are associated with lime- from observations of exposures in twenty seven inliers, stone and calcareous shale containing the rudist bivalve together with information from nine onshore and two off- Barrettia monilifera and the larger benthic foraminifer shore exploration wells. Figure 6.4 is a simplified correla- Pseudorbitoides trechmanni, which elsewhere occur with tion diagram for the Cretaceous rocks of the major inliers. Middle Campanian calcareous nannofossils 58. These rocks Correlation is complicated by the variations in the largely are overlain by the Belleview Group, a thick sequence of volcanogenic facies. volcanogenic sedimentary rocks, associated with flows and pyroclastics, in turn overlain by the shallow -water, rudist- Blue Mountains Block bearing carbonates of the Rio Grande Limestone (uppermost The Blue Mountains Block extends over much of the Campanian? to Lower Maastrichtian). eastern third of Jamaica, with about 500 square km of In the Sunning Hill district, calcalkaline volcanogenic Cretaceous rocks being exposed in the Blue Mountains- sedimentary units similar to those of the Back Rio Grande Sunning Hill Inlier. The block is faulted along most of its and Belleview Groups also contain B. monilifera, together eastern, southern and western margins, but, along most of with middle Campanian inoceramid bivalves109 (Bon Hill the northern flank, Tertiary strata dip northwards off the and Thornton Formations). In contrast, the ophiolitic suite Cretaceous rocks of the Inlier. in the southeastern Blue Mountains consists of a basal Overall the structure of the block is antiformal with complex of pillowed tholeiitic basalt, gabbro and radiolarian major vertical displacement along the Plantain Garden chert, overlain by mudstone and cherty limestone with Fault associated with an outcropping belt of metamorphic planktic foraminifers and reworked larger foraminifers. rocks along the southern edge of the inlier (Figs 6.1, 6.2). This was deposited during the Campanian to early Maas- The maximum positive bouguer gravity anomaly exceeds trichtian, but in bathyal to abyssal environments, possibly 155 mgal, occurring over blueschists, part of the dense, at as much as 4 km water depth55 (Bath-Dunrobin Complex; metamorphic belt. A two dimensional model, proposed to Fig. 6.4). explain the gravity anomaly, emphasizes the vertical dis - 108 Volcaniclastic shales and sandstones, at least partly placement on the Plaintain Garden Fault . The structu- turbiditic in origin (Cross Pass Shales), apparently overlie

113 Jamaica

Figure 6.2. Simplified structural map and rose diagram of fault traces of Jamaica (modified from Draper30). Stippled areas are blocks, unornamented areas are belts. Numbered fault systems are; l=Spur Tree; 2=Crawle River, 3=Duanvale; 4=Wagwater; 5=South Coast; 6=Plantain Garden. The rose diagram is of fault traces from the geological map of McFarlane74.

both the calcalkaline and ophiolitic suites, and are succeeded synsedimentary faults. The Clarendon Block is distin- gradationally by massive polymict conglomerates (Bowden guished morphologically by extensive upland areas, cov- Pen Conglomerate). The ages of these formations are based ered by rugged karst terrain, but numerous faults have on the rare occurrence of Pseudorbitoides sp. and double- broken up the landscape. Several, such as the Spur Tree keeled Globotruncana sp. (Campanian-Maastrichtian), and Fault, have produced spectacular escarpments and record on nannofossils of similar antiquity. The Bowden Pen Con- relatively young movements, post-dating the formation of glomerate includes fragments reworked from the Rio the major bauxite deposits, developed on Miocene lime- 64 Grande Limestone or its equivalent . Both the Cross Pass stones in the area87. Shales and Bowden Pen Conglomerate are in mainly faulted A number of Cretaceous inliers (Fig. 6.3) expose rocks contact with Paleocene shales of the John Crow Mountains ranging in age from pre-Barremian to Maastrichtian, to- 58 Belt (Moore Town shales ). The Clarkes River Formation gether with a single outcrop, at the west side of Kingston of the Sunning Hill district (Fig. 6.4) is probably equivalent Harbour (Green Bay Schists; Fig. 6.4), of amphibolites, to the Cross Pass Shales, Well-carbonized phytoclastic ma- identical to the Westphalia Schist of the southwestern Blue 109 terial occurs within both units . Mountains62. Granitoid intrusives occurring in the Blue Mountain The Benbow Inlier (Fig. 6.3) contains the oldest, pa- 54 Inlier have been discussed by Isaacs and Jackson , who laeontologically-dated rocks on the island (early Creta- showed that the easternmost plutons have tholeiitic affini- ceous), but older rocks may occur in the Above Rocks Inlier, 110 ties, and are apparently the oldest (80 ± 5 Ma ), while the to the south. In the Above Rocks Inlier calcalkaline grani- others are calcalkaline. toids, dated isotopically at 63 ± 3 Ma19,42, are intrusive into undated, northward-dipping, volcanogenic and siliceous Clarendon Block 83 104 sedimentary rocks, now hornfelsed . The Clarendon Block was named by Versey , who Within the Benbow Inlier Burke et al.11 distinguished introduced the concept of blocks and belts, separated by some 4000 m of volcanic flows, volcanogenic conglomer-

114 EDWARD ROBINSON

Figure 6.3. Major Cretaceous inliers, Cretaceous and Cenozoic volcanic centres and petroleum exploration wells of Jamaica. Key to Cretaceous inliers: l=Lucea; 2=Jerusalem Mountain; 3=Grange; 4=Marchmont; 5=Mocho; 6=Maldon; 7=Sunderland; 8=Nottingham; 9=Central; 10=St Anns Great River; ll=Benbow; 12=Above Rocks; 13=Gibla- tore; 14=Blue Mountain. Key to volcanic centres: black spots=Cretaceous volcanic centres; stars=Paleocene-Eocene dacite centres (NE is Newcastle); crosses=Paleocene-Eocene basalt centres (HA is Halberstadt, NU is Nutfield); triangles=late Neo- gene basalt centres (LO is Low Layton, MO is offshore Morant Point): Key to sites of petroleum exploration wells, open circles: A=Pedro Bank (inset); B=Arawak (inset); C=west Ne- gril; D=Negril Spots; E=Hertford; F=Content; G=Retrieve; H=Cockpit; J=Santa Cruz; K=Portland Ridge; L=Windsor. Other localities: RH=Round Hill; KN=Kingston; MB=Montego Bay; SU=Sunning Hill. Key to geological ornament: diagonal lines=formations of Coastal Group; Wawfc=White Limestone Supergroup and Yellow Limestone Group; cross-hatch=early Tertiary rocks of Wagwater Belt; stipple=Cretaceous rocks.

ates, sandstones and shales, and rudist limestones, which turbiditic and poor in fossils, but include Turonian calcare- extend from the Lower (pre-Barremian) to Upper Creta- ous nannofossil assemblages 58 (Rio Nuevo Formation). ceous (Turonian). The oldest rocks (Devils Race Course The 300 square km of Cretaceous rocks in the Central Group) are hydrothermally altered andesitic lavas, suc - Inlier (Fig. 6.3) form an east-west trending antiformal struc- ceeded by mixed volcanogenic rocks and shallow water ture cut by axially situated sinistral wrench faults (Crawle limestones (pre-Barremian to Aptian; Copper, Jubilee and River Fault System; Fig. 6.2). The exposed sequence ranges Benbow Limestone Members). These are overlain by pillow from Santonian or older to Paleocene, overlain with angular lavas of island arc tholeiitic affinity55 and the Albian unconformity by an early Tertiary transitional to shallow Seafield Limestone Member. This succession is overlain by marine clastic and carbonate succession. volcanogenic sandstones and shales. These are commonly Geographically and stratigraphically there is a broad

115 Jamaica

division into a middle Cretaceous, volcanogenic, possibly Hill Formation conformably, the latter unit having locally hydrothermally altered province, concentrated in the eastern suffered contact metamorphism. and central parts of the inlier, and a Campanian to Maas - The western part of the inlier exposes basal redbeds and trichtian/Paleocene, clastic/carbonate sequence in the cen- volcanogenic siltstones (Slippery Rock Formation), uncon- tral and western parts of the Inlier, unaffected by contact or formable on the Bullhead/Main Ridge Volcanics, and pass- hydrothermal metamorphism. General accounts of these ing up into the Guinea Corn Formation, locally crowded rocks have been written by Porter 79 (eastern end), Coates21 with rudists, including very large forms, such as Titanosar- (central portion), Robinson and Lewis 93, Kauffman and colites giganteus. The marine part of this section is Upper 58 Sohl61 flate Cretaceous carbonates of the western end), and Campanian to Lower Maastrichtian . Roobol95 (Upper Cretaceous-Paleocene volcanics). The The overlying Summerfield Formation interdigitates presence of pervasive low grade metamorphism in Jamaican with the top of the Guinea Corn Formation and marks a Cretaceous rocks in general was discussed by Meloche75. reversion to volcanic activity in the region. Formerly re- The eastern volcanics consist of poorly-stratified vol- garded as uppermost Cretaceous, this unit is now thought to caniclastic rocks associated with amygdaloidal basalts, por- be at least partly of Paleocene to early Eocene age (see phyritic basaltic andesites and flow-banded lavas, intruded below). by a granodiorite stock. Flanking the stock to the southwest The Sunderland, Marchmont and most of the other, is a 200 m wide belt of 'porphyry-copper type' altered rocks. smaller inliers in the western part of the Clarendon Block Contact metamorphism has produced a narrow, but well-de- lack the altered volcanic suites of the Central Inlier, and fined, aureole of pyroxene and hornblende hornfels. The contain only Santonian to Maastrichtian sedimentary se- whole sequence is invaded by mafic intrusives as a dyke quences. In the Sunderland Inlier, the lowest horizons are swarm. The 1000+ m thick Arthur's Seat Formation to the of conglomerates and shales (Johns Hall, Sunderland and west is also dominated by poorly-bedded, unsorted, vol- Newmans Hall Formations), of late Santonian to middle 58 canic (laharic to epiclastic) conglomerates and breccias Campanian age , succeeded by Barettia-bearing limestone associated with subordinate laminated, feldspathic, sand- (Stapleton Formation), redbeds and Titanosarcolites stones and shales, and, more rarely, lava flows. Low grade limestones similar to those of the Central Inlier (Shepherds (zeolite facies) metamorphism has affected the entire forma- Hall to Vaughnsfield Formations; Fig. 6.4). In the tion. The granodiorite intrusion has been isotopically dated Marchmont Inlier only the higher part of the sequence is at 72.7 ±2.0 Ma67. present. Weathered tuffs and possible flows occur only in Epiclastic volcanics flanking the area of the stock are the Nottingham Inlier (Fig. 6.3), resembling the lithology of overlain by rudist-bearing limestone, siltstone and shale the Arthur's Seat or Main Rid ge/Bullhead Formations. Else- with a fauna including oysters, the rudist Barrettia, corals where along the south side of the Cockpit Country minor and the inoceramid Platyceramus (I.) cycloides, indicative exposures of Summerfield Formation-like lithology are of the Middle Santonian to Middle Campanian interval. seen. Immediately to the west, the 400 m thick Peters Hill Forma- Off the northern edge of the Clarendon Block, in the tion is lithologically similar, consisting of mudstone and North Coast Belt, the small St. Ann's Great River Inlier siltstone with subordinate sandstone, and with a conspicu- (Figs 6.3, 6.4) exposes a section of volcanogenic rocks of ous limestone at the base (Fig. 6.4). This contains a rich Coniacian to Campanian age76. fauna of solitary and compound corals, nerineid gastropods and rudist bivalves. This unit is Middle to Upper Santonian, Hanover Block based on species Inoceramus and calcareous nannofossils in Four inliers in the Hanover Block contain Santonian to the shales58,60. Maastrichtian sedimentary rocks. In the largest, the Lucea In the central part of the inlier the Bullhead/Main Ridge Inlier, Grippi40 distinguished fourteen units in three small Volcanics consist of poorly-bedded and poorly-sorted feld- blocks, separated by two east-west faults. In the largest, spathic grits, sandstones, conglomerates and breccias, asso- central block, a 4000 m sequence of volcaniclastic sand- ciated with porphyritic andesite and basalt, and aphyric stones and shales ranges in age from Santonian to Middle basalt flows. Total thickness is variable between 750 and Campanian, and includes a minor limestone (Clifton Lime- 1000 m, and the sequence has been intruded by numerous stone). The smaller northern block exposes Campanian vol- dykes of aphyric and porphyritic hornblende basalt, basaltic caniclastics, while the southern block contains andesite and pyroxene-rich andesite. There is extensive, Santonian-age sandstones. Part of the sequence in the inlier though variable, low grade (zeolitic) hydrothermal altera- has been interpreted as being formed within a submarine tion. Although contacts are mainly faulted, the Bull- canyon complex with conglomerate channel fill cutting head/Main Ridge Volcanics apparently overlie the Peters across the main underlying sequence41. Schmidt's97 revi-

116 EDWARD ROBINSON

sion of the geology reduced the number of recognized by limestones, important terrigenous clastic sedimentation formations to six (Fig. 6.4) and suggested that they were occurred in the Paleocene to middle Eocene, and again from deposited as a submarine fan complex. late Miocene times onward (Figs 6.5,6.6). The other inliers are much smaller, and collectively contain a succession similar to, but thicker than that of, the John Crow Mountains Belt western part of the Clarendon Block. Campanian marine Paleocene marine strata are confined to the region east of shales (Dias Formation), followed by a locally developed the Wagwater Fault (Fig. 6.2). The thickest marine se- limestone horizon with the rudists Barrettia and Paras- quence, dated by planktic foraminifera and calcareous troma (Green Island Formation), pass up through redbeds nannofossils 58,98, occurs in the John Crow Mountains Belt (Morelands beds) into shallow-water, rudist limestones with (Fig. 6.2), where Lower Paleocene shales and subordinate Titanosarcolites (Thicket River and Jerusalem limestones). turbiditic sandstones and conglomerates, up to 4000 m thick, Overlying the limestones is a redbed sequence (Masemure were deposited, at least in part, at bathyal depths (Moore Formation), stratigraphically the equivalent of the Summer - Town shales). They are succeeded by Upper Paleocene, field Formation of central Jamaica76. impure bioclastic (algal) and micritic limestones and calcareous mudstones, also at least partly bathyal (Nonsuch Subsurface Cretaceous Rocks Limestone). Detailed field relationships are still obscure, Of the eleven wells drilled for petroleum onshore and with mainly faulted contacts, but the Moore Town shales are offshore (Fig. 6.3), eight penetrated rocks of Cretaceous age, underlain by Upper Cretaceous to possible Paleocene con- 58,64,102 three in the Negril - Savanna-la-Mar Belt34, three on the glomerates (Bowden Pen Formation) and shales Clarendon Block, one in the North Coast Belt, and one on (Cross Pass Shales and Providence Shales). It is likely that Pedro Bank76,82. In the Negril - Savanna-la-Mar Belt, the basinal sedimentation was more or less continuous from the Negril Spots-1 and West Negril-1 encountered an Upper Cretaceous to the Paleocene. The Paleocene rocks of the John Cretaceous volcaniclastic section below Tertiary lime- Crow Mountains Belt are overlain unconformably by a 58 stones. The Hertford-1 well reached Santonian volcaniclas- middle Tertiary carbonate sequence (White Limestone tics82. Supergroup). On the Clarendon Block, Cretaceous rocks in the Cock- pit-1 well are similar to those outcropping in the western Blue Mountains Block part of the Central Inlier. Retrieve-1 was drilled entirely in Upper Paleocene shallow water carbonates (Chepstow Limestone) border the northern and western margins of the Cretaceous volcaniclastics down to a possible Turonian age 92 horizon82. Santa Cruz-1 drilled through a thick Tertiary Cretaceous Blue Mountain Inlier . At Chepstow this unit carbonate sequence (White and Yellow Limestone Groups) overlies Upper Cretaceous volcanics and granodiorite. The to penetrate, at 1966 m, a green to grey and brown siliceous, Chepstow Limestone is succeeded by Lower Eocene silici- conglomeratic sandstone, 18 m thick. Drilling continued clastics (Richmond Formation) and Middle Eocene impure through 425 m of andesitic flows and ?sills or dykes, meta- limestone (Yellow Limestone Group) described more fully morphosed to zeolite facies. Below this, to total depth (2,662 below. m), metamorphosed amphibolitic rocks were reported. Wagwater Belt These rocks may correspond to some part of the Arthur's Seat Formation or the older volcanics from the east end of The early Tertiary succession in the Wagwater Belt the Central Inlier76. consists of at least 5.6 km of coarse-grained, terrigenous, In the St Ann's Great River Inlier, the Windsor-1 well dominantly terrestrial strata, confined to a relatively narrow spudded in Coniacian strata and drilled through a volcani- zone in the belt, overlain by about 1.2 km of upward-fining clastic section into Lower Cretaceous rocks similar to those marine rocks occupying a much larger region, encompass- of the northern part of the Benbow Inlier82. Offshore, Pedro ing the northern, western and southern flanks of the Blue Mountains Block, as well as the Wagwater Belt itself70. Bank-1 drilled a Tertiary carbonate sequence underlain by 39,70 granitoid rocks similar to those of the Clarendon Block76. The lower, coarse-grained unit (Wagwater Forma- tion) is dominated by redbeds and contains evaporites in its Arawak-1 penetrated a much thicker section of Tertiary 5,47 carbonates59. lower part (Brooks Gypsum), but also includes minor limestones with oysters, small corals, ostracodes and algae (Woodford Limestone of Chubb17; Halberstadt Limestone EARLY TERTIARY 73 of Matley ). Its Paleocene age is inferred by its strati- Although the Tertiary succession of Jamaica is dominated graphic position below beds in the upper part of the formation, correlated with Upper Paleocene beds elsewhere in

117 Jamaica

Figure 6.4. Correlation of main Cretaceous to ?Lower Paleocene lithostratigraphic units of Jamaica.

eastern Jamaica6. The Wagwater Belt has been interpreted Over the region of the Wagwater Belt and Blue Moun- as representing deposition within a fan delta and proximal tains Block the siliciclastics of the Richmond Formation are submarine fan complex111. succeeded by impure, bioclastic strata and planktonic The upper, laterally more extensive, Richmond Forma- foraminiferal marls of the Font Hill Formation (Yellow tion contains marine body and trace fossils indicative of an Limestone Group). In the northern part of the Wagwater Belt early Eocene age and deposition at sublittoral to bathyal the conformable transition from the Richmond Formation to depths 58,77. Westcott and Etheridge111 interpreted the unit the Font Hill Formation occurs just above the base of the as representing accumulation mainly as distal submarine fan Middle Eocene, the foraminiferal faunas of the latter unit deposits. Mann and Burke70 have recognised a number of being indicative of deposition in outer sublittoral to bathyal smaller units, of member rank, within the Wagwater and palaeodepths. In the southern Wagwater Belt the Font Hill Richmond Formations, in addition to those indicated in Formation occurs as mixed planktic foraminiferal marls, Figure 6.5 herein. bioclastic turbidites, and slumped and olistostromic units of The Wagwater Formation is intercalated with mainly Middle Eocene age12. In the central part of the Wagwater extrusive, calcalkaline, dacitic flows of the Newcastle Vol- Belt, and over most of the Blue Mountains Block, beds of canics from, perhaps, three centres within the Wagwater this age are missing, presumably because of subsequent Belt56 (Figs 6.3, 6.5). Isotopic dates range in the interval erosion. On the north flank of the Blue Mountains the 45-35 Ma2, but stratigraphic considerations favour a Yellow Limestone Group is represented by shallow water Paleocene to possibly early Eocene age70,92. carbonates, similar to those of central Jamaica92. Other volcanic rocks occur as flows from local centres (Fig. 6.3). In the lower part of the Wagwater Formation the Central and Western Jamaica Halberstadt tholeiitic pillow basalts were probably extruded In central and western Jamaica volcanigenic, pyroclas- in water depths of about 300 m55,94. At the northern end of tic and epiclastic strata (Figs 6.4,6.5) overlie Maastrichtian the Wagwater Belt pillow basalts and dacites of the Nutfield shallow marine carbonates. The Summerfield Formation Volcanics were intruded and extruded as part of the Rich- consists of grey- to lilac-coloured volcaniclastics, including mond Formation94 (Fig. 6.5). prominent pumiceous layers, and two ignimbrite hori-

118 EDWARD ROBINSON

zons21,95. The monotonous mineralogy and presence of face. The name White Limestone was first used by de la pumice throughout the Summerfield Formation suggests Beche7 to describe all the Tertiary limestones of Jamaica, that it represents a single volcanic episode95. Allen pro- but the name is now restricted for the nearly pure, middle posed a threefold division of the formation in the northwest- Tertiary carbonates which everywhere follow on top of the ern part of the Central Inlier, the highest of which is probably Yellow Limestone Group. Matley72'73 attempted to zone the equivalent to the basal, quartz-rich, terrigenous sedimentary White Limestone using larger foraminifers, his efforts being strata of the overlying Eocene Chapelton Formation. A followed by a more comprehensive scheme developed by fission track date of 55.3 ±2.8 Ma for apatites from a sample Hose and Versey53 and Versey1 4. The latter writers also of Roobol*s upper ignimbrite3 indicates an early Tertiary attempted a subdivision of the White Limestone into a (Paleocene to early Eocene) age for the Formation, rather number of smaller lithological units, based on microfacies than the late Cretaceous age previously accepted76,117. In analyses of the limestones. Field mapping and elaboration of other parts of central and western Jamaica ignimbrites are these units was achieved over much of the island in the lacking and the redbed equivalents of the Summerfield period 1955 to 197574. Formation (Masemure, Garlands-Mocho beds; Fig. 6.4) are Although the microfacies units, recognized by Hose more epiclastic in appearance. and Versey as members of their White Limestone Forma- These end-Cretaceous to Paleocene volcanigenics and tion, have often been given formational status89'114, their redbeds are followed unconformably by interbedded, im- recognition is mainly based on a combination of lithostrati- pure limestones, sandstones and mudstones, typically rich graphic and biostratigraphic criteria. In this chapter the in shallow marine macrofossils and with minor beds of White Limestone Supergroup is divided into two distinct lignite (Chapelton Formation; Fig. 6.5). Around the Central lithostratigraphic units, the Montpelier and Moneague Inlier a fourfold division of the formation is distinguished. Groups 45. The smaller units of Hose and Versey, and others, Thin, commonly discontinuous, basal quartz-rich sand- although not all lithostratigraphic units in the strict sense of stones and conglomerates (Freemans Hall Beds) are suc- the definition112, are regarded here as formations, because it ceeded in the north by the richly fossiliferous limestone of has proved possible to map them over large parts of the Stettin Member, of late early Eocene age85. The Stettin Jamaica74, and their names are well entrenched in the litera- Member is overlain by a terrigenous unit (Guys Hill Mem- ture34,76,114,117. ber) of sandstones, mudstones and minor, impure lenticular Along the north coast, over much of the area flanking limestones. The carbonates are mainly biostromal oyster the Blue Mountains, around the edge of the Hanover Block, and Carolia beds, but include at least one lens yielding and in the Montpelier-Newmarket Belt, the Montpelier crocodilian and sirenian remains 8,27. Discontinuous beds of Group occurs as a series of evenly bedded chalks or micrites, lignite occur in the middle of the Guys Hill Member88. characterized by interbedded nodular or platy layers of chert in Where the Stettin Member is absent, as on the southern flank all but the highest part (Fig. 6.6). Fossil assemblages are of the Central Inlier, and around the Benbow Inlier, the dominated by planktic foraminifers representing deposition at 99,100 Freemans Hall beds lose their identity, and the Guys Hill lower sublittoral to abyssal depths . Member forms the basal unit of the Chapelton Formation. Over the Clarendon and Hanover Blocks, and fringing In the area around Kingston, a thin equivalent of the Guys parts of the Blue Mountain Block, massive to thick-bedded, Hill Member (White Limestone Basement of Matley73) coral, algal, molluscan, echinoidal and benthic foraminiferal contains gypsiferous beds, including localities yielding gyp- limestones of the Moneague Group occur. These rocks may sum pseudomorphs after halite, and layers interpreted as be patchily siliceous, but lack bedded chert and some are indicative of a sabkha depositional setting13. These beds dolomitic. The lower parts of the unit are generally well- directly underlie the White Limestone Supergroup. lithified, while the higher parts may be rubbly to earthy in The upper part of the Chapelton Formation consists of texture. Fossils may be locally abundant, but are commonly impure limestones yielding a diverse biota of molluscs, poorly preserved. The assemblages are characteristic of echinoids, algae and foraminifers, of Middle Eocene age deposition at sublittoral depths within the photic zone. (Albert Town Member). Towards the western part of the The two groups are lateral equivalents, reflecting con- island impure carbonates appear to comprise the major part temporaneous deposition on shallow carbonate banks of the Chapelton Formation and subdivision of the unit has (Moneague Group) and as periplatform accumulations on not been attempted in that area. the flanking, deep sea floor (Montpelier Group). In Fig. 6.6 the main formations and microfacies units are indicated and White Limestone Supergroup briefly described. The palaeoenvironmental interpretation More or less pure carbonates of the White Limestone and distribution of the units have been discussed by 104 99 34 Supergroup89 outcrop over more than half the island's sur- Versey , Steineck , and Eva and McFarlane .

119 Jamaica

Figure 6.5. Correlation of main Lower Tertiary lithostratigraphic units of Jamaica.

LATE TERTIARY AND QUATERNARY lower Coastal Group consists mainly of planktic foraminiferal marls86 (Buff Bay and Bowden Formations, and San Middle Miocene to Quaternary sedimentary rocks of Ja- San Beds) deposited at outer sublittoral to bathyal depths99. maica form the Coastal Group. They outcrop as discontinu- The marine Quaternary rocks of the upper Coastal ous strata along the margins of the island and incorporate a Group consist of a lower unit (Manchioneal Formation) of wide range of lithologies. The lower, pre-Quatemary part of foraminiferal and coral/mollusc-rich marls, locally with a the Coastal Group, like the underlying White Limestone significant brachiopod fauna43,86,103, found mainly along Supergroup, is divisible into two broad units, representing the northeastern coast of the island. The dolomitised coral deposition in contrasting paleoenvironments (Fig. 6.6). 18,7 3 limestone of the Hope Gate Formation of the north coast is Along the south coast, just east of Kingston and at 65 28 probably of similar age , as are the marginal marine, mixed Round Hill, parish of Clarendon (Fig. 6.3), a mixed se- siliciclastics and marls of the Harbour View Formation near quence of terrigenous conglomerates and sandstones, im- Kingston9. The higher part of the upper Coastal Group is pure limestones, coral-block conglomerates, marls and represented principally as a series of raised marine terrace calcareous siltstones constitute the August Town Forma- deposits. In places the terrace sequences reach altitudes of tion. This unit contains a variety of fossils of shallow marine 200 m15. The lowest terrace group, containing the Falmouth derivation, from pebble beaches, coral reefs and oyster Formation (Fig. 6.6), has been dated at 0.13 Ma15. Within banks. Although probably deposited at shelf depths for the the terrace containing the Falmouth Formation, facies in- most part, similar modern, coarse-grained clastic sediments, dicative of reef and lagoonal depositional environments evidently redeposited from the rivers and beaches adjacent have been distinguished at several localities 15,35,66,84. to Kingston, have been dredged from bathyal depths south 10 Non-marine Quaternary deposits include extensive de- of Kingston (Robinson, unpublished observations). In velopment of alluvium across the coastal plain along the contrast, along the north and east coasts of Jamaica, the southern margin of the island, alluvium related to the poljes

120 EDWARD ROBINSON

of the karst limestone areas, and fanglomerates associated Paleocene and extending to form the Montpelier-Newmar- 33,69 with fault scarps. The largest of these is the fan (Liguanea ket Belt by the late middle Eocene . These movements Formation) on which the city of Kingston is built36,73. The produced the block and belt features (Fig. 6.2) which were bauxite deposits of the interior are mainly confined to areas to control sedimentation patterns over the island for the underlain by the shallow water carbonate units of the White remainder of the Tertiary. Limestone Supergroup (Moneague Group of this chapter). Redbeds, associated with plateau type tholeiites, eva- The relationship of these deposits to the bedrock has been porites and local very shallow marine incursions charac- the subject of numerous papers (for example, in Lyew- terized the early development of the narrow graben of the 70 Ayee68). Wagwater Belt . In the John Crow Mountains Belt, early Paleocene, deep marine, flysch sedimentation, possibly continuous from the Cretaceous, resembles sedimentary pat- TECTONIC EVOLUTION terns seen in southwestern Haiti105 and southern Belize25 (Sepur Group). Over the remainder of the island this transi- The tectonic evolution of Jamaica is conveniently divided tional period was marked by subaenal andesitic volcanic 31 into four phases, following Draper ; a Cretaceous island activity (Summerfield Formation and Masemure Beds). arc phase, a Paleocene to middle Eocene transitional phase, a During the late Paleocene, a more extensive shallow middle Tertiary period of relative tectonic quiescence, and a marine transgressive phase produced the carbonates (Chep- late Neogene to Recent phase of renewed tectonic activity. stow Limestone) fringing the Blue Mountains Block and similar, but deeper-water, depositional systems extended Cretaceous Island Arc over the John Crow Mountains Belt (Nonsuch Limestone). It is generally agreed that subduction, with a west In the Wagwater Belt extrusion of the Newcastle Volcanics dipping component, was taking place throughout the Creta- was taking place. ceous, associated with a volcanic arc, of which Jamaica By the early Eocene the narrow, Wagwater graben had 30,31,52,78 formed a part . In Jamaica the arc rocks are represented evolved into a broader area of siliciclastic deposition70 by the calcalkaline volcanogenic assemblages, while (Richmond Formation). This area encompassed much of subduction is indicated by the amphibolites of the metamor- eastern Jamaica (except, possibly, the John Crow Mountains 31 phic complex of the Blue Mountains Inlier . Belt, where early Eocene strata have not yet been identified) 57 Jackson et al. showed that geochemical variations and the newly forming North Coast Belt. Palaeocurrent among statistically selected major, minor and trace elements directions are either transverse to or parallel to the axis of in Upper Cretaceous volcanic and plutonic rocks could be the Wagwater Belt14. By late early Eocene times marine interpreted as cross-arc trends and used to infer differences transgression extended over parts of central and western in palaeo-crustal thicknesses over a subduction zone. The Jamaica (Stettin Limestone). trends suggest that the volcanics of eastern Jamaica (Blue Several alternative mechanisms have been suggested to Mountains Inlier) formed over a shallower subduction zone explain the changes in tectonic style at this time. They and thinner (25-30 km depth) crust than volcanics of similar include back-arc basin development in response to west or age and composition further west. As post-Cretaceous tec - southwest directed subduction56, localized extension result- tonic rotation of Jamaica is believed to have amounted to ing from movement of the Nicaraguan Rise past the Yucatan 37,38 only some 10° of counterclockwise motion , these con- block of Central America69, and pull-apart basin formation clusions lend support to tectonic models postulating a west during a brief period of right-lateral shear in the Nic araguan dipping component in a late Cretaceous subduction zone. Rise region30. Draper30 proposed a southeastwards shift in the position of the subduction zone during the Cretaceous, to ac- Middle Tertiary Quiescence count for the juxtaposition of the early Cretaceous Between the middle Eocene and middle Miocene, dif- subduction complex metamorphics of southeast Jamaica ferential subsidence allowed as much as 2.75 km of exclu- against the late Cretaceous volcanic arc suite. sively carbonate platform deposits to accumulate over the blocks31. Probably no part of the island region was more Early Tertiary Transition than a few metres above sea level at any time. The structural The mechanism for the transition from Cretaceous sub- trends developed in the early Tertiary, and bounding the duction to early Tertiary horst and graben formation is still blocks and belts, now served to maintain topographic dif- not fully understood. This early Tertiary phase was marked ferences between shallow water platform deposition on the by northwest-southeast trending fracturing of the Jamaican blocks and periplatform accumulation in the basins between region, initiating formation of the Wagwater Belt in the the blocks34. Within these basins, particularly those fronting

121 Jamaica

Figure 6,6. Correlation of Middle Eocene to Quaternary lithostratigraphic units of Jamaic a.

the Cayman Trough, water depths reached as much as 2 Late Cenozoic uplift of Jamaica accompanied this tec- km99,100. During the early and middle Miocene, southward tonic activity and was particularly intensive in the east, with tilting of the Jamaican region is implied by the southward late Miocene Coastal Group strata in the northeast being thickening, wedge-shaped geometry of the Newport Forma- elevated as much as 2000 m100 and early Pleistocene units tion113. raised some 200 m or more. Late Pleistocene raised marine terraces, mainly along the north coast, document differential Late Cenozoic Activity uplift and faulting, continuing to the Holocene15,48,50. The Evidence for the unroofing of parts of the Jamaican Holocene rise in sea level drowned parts of the island's middle Tertiary carbonate platform is found in the non-car - margins, leading to the development of the modem wetlands bonate detritus shed into the Coastal Group from the end of of the south and west coastal areas 44. the middle Miocene87. Activity on east-west, left lateral Indirect evidence for volcanic activity in the region transcurrent faults, northeast-southwest normal faults, and surrounding Jamaica, especially in the late Cenozoic, has northwest-southeast trending reverse and scissors faults, been proposed through the suggestion that much, if not and folds, which dominate the present structural pattern of most, of the superficial cover of bauxite over the limestones Jamaica, has been discussed by Draper31. The Wagwater of central Jamaica is derived from weathering of Tertiary Fault, intiated as a normal fault in the Paleocene, reversed ashfall deposits, some of which are still preserved as ash its movement to become a high-angle thrust fault71, with beds in the deep water limestones of the Montpelier associated deformation including local overturning of the Group23,24. Pliocene August Town Formation.

122 EDWARD ROBINSON

ACKNOWLEDGEMENTS—This chapter includes the results of numer- Eocene. Journal of the Geological Society of Jamaica, ous discussions over thirty years with the geological frat ernity of Jamaica, 12,1-6. past and present. I do not wish to make special mention of any one or more 14 persons who have provided ideas and discussed data with me, but the names Cambray, F.W. & Jung, P. 1970. Provenance of the Rich- of most of them are to be found in the list of references below. I also thank mond Formation from sole marks. Journal of the two anonymous reviewers of this chapter for their instructive comments. Geological Society of Jamaica, 11,13-18. 15Cant, R.V. 1972. Jamaica's Pleistocene reef terraces. Journal of the Geological Society of Jamaica, 12, 13- REFERENCES 17. 16 1 Case, J.E. & Holcombe, T.L. 1980. Geologic -tectonic Ahmad, R. (ed.). 1987. Proceedings of a Workshop on the map of the Caribbean region. U.S. Geological Survey Status of Jamaican Geology. Geological Society of Miscellaneous Investigations Series, Map I-1100. Jamaica, Special Issue, 342 pp. 17 2 Chubb, L.J. 1956. A subsidence in the mountains of Ahmad, R. & Jackson, T.A. 1989. Pre-Quatemary geo- Jamaica. Colonial Geology and Mineral Resources, 3, chronological studies in Jamaica: a review (abstract). 127-132. Journal of the Geological Society of Jamaica, 25,38. 18Chubb, L.J. 1958. The higher Miocene of south-eastern 3Ahmad, R., Lai, N. & Sharma, P.K. 1988. Fission-track age Jamaica. Geonotes, 2, 26-31. of ignimbrite from Summerfield Formation, Jamaica 19Chubb, L.J. & Burke, K. 1963. Age of the Jamaica grano- Caribbean Journal of Science, 23, 444-448. 4 diorite. Geological Magazine, 100, 524-532. Allen, L., 1985. Lithofacies of the Summerfield Formation, 20CIDA/GOJ. [Siriunas, J.M.] 1992. Geochemical map of northwestern Central Inlier, Jamaica. Abstracts, Work - Jamaica. CIDA/GSD no. 504/12713-142061 Metallic shop on Recent Advances in Jamaican Geology, Geo- Mineral Survey of Jamaica, Phase II, Open File Report, logical Society of Jamaica, Kingston, 2. 5 1. Allen, L. & Neita, M. 1987. Geology of the north Bull Bay 21 Coates, A.G. 1968. The geology of the Cretaceous Central sulphate occurrence zone: in Ahmad, R. (ed.), Proceed- Inlier around Arthur's Seat, Clarendon, Jamaica: in ings of a Workshop on the Status of Jamaican Geology. Saunders, J.B. (ed.), Transactions of the 4th Caribbean Geological Society of Jamaica, Special Issue, 282-298. 6 Geological Conference, Port-of-Spain, Trinidad, 28th Armour-Brown, A. 1967. Buff Bay Quadrangle. Annual March-12th April, 1965, 309-315. Report of the Geological Survey Department, Jamaica, 22Coates, A.G. 1977. Jamaican coral-rudist frameworks and for 1966,9-10. 7Beche, H.T. de la. 1827. Remarks on the geology of Ja- their geological setting: in Frost, S.H., Weiss, M.P. & maica. Transactions of the Geological Society of Lon- Saunders, J.B. (eds), Reefs and Related Carbonates in Ecology and Sedimentology. American Association of don, series 2,2, 143-194. 8 Petroleum Geologists, Studies in Geology, 4, 83-91. Berg, D.E. von. 1969. Charactosuchus kugleri, eine neue 23 Krokodilart aus dem Eozan von Jamaica. Eclogae Comer, J.B. 1974. Genesis of Jamaican bauxite. Economic Geology, 69,1251-1264. Geologicae Helvetiae, 62, 731-735. 24 9Bold, W. A. van den. 1971. Ostracoda of the Coastal Group Comer, J.B., Naeser, C.W. & McDowell, F.W. 1980. of formations of Jamaica. Transactions of the Gulf Fission-track ages of zircon from Jamaican bauxite and terra rossa. Economic Geology, 75, 117-121. Coast Association of Geological Societies, 21, 325- 25 348. Donnelly, T.W., Home, G.S., Finch, R.C. & Lopez-Ra- 10Burke, K. 1966. The Yallahs Basin: a sedimentary basin mos, E. 1990. Northern Central America; the Maya and southeast of Kingston, Jamaica. Marine Geology, 5, Chortis blocks: in Dengo, G. & Case, J.E. (eds), The 45-60. Geology of North America, Volume H, The 11Burke, K., Coates, A.G. & Robinson, E., 1968. Caribbean Region, 37-76. Geological Society of America, Boulder, Colorado. Geology of the Benbow Inlier and surrounding areas: 26 in Saunders, J.B. (ed.), Transactions of the 4th Donovan, S.K. 1993. Geological excursion guide 9: Ja- maica. Geology Today, 9,30-34. Caribbean Geological Conference, Port -of-Spain, 27 Trinidad, 28th March-12th April, 1965, 299-307. Donovan, S.K., Domning, D.P., Garcia, F.A. & Dixon, 12Burke, K. & Robinson, E. 1965. Sedimentary structures H.L. 1990. A bone bed in the Eocene of Jamaica. Journal of Paleontology, 64,660-662. in the Wagwater Belt, eastern Jamaica. Journal of the 28 Geological Society of Jamaica, 7,1-10. Donovan, S.K., Jackson, T.A. & Littlewood, D.T.J. 1989. 13Burne, R.V. 1972. Supra-tidal evaporites from the Grants Report of a field meeting to the Round Hill region of Pen Clay: evidence of drier conditions in the Jamaican southern Clarendon, 9 April, 1988. Journal of the Geo- logical Society of Jamaica, 25, 44-47.

123 Jamaica

29Draper, G. 1986. Blueschists and associated rocks in The Phanerozoic Time Scale, a Symposium. Geological eastern Jamaica and their significance for Cretaceous Society of London Special Publication, 1, 458 pp. 43 plate-margin development in the northern Caribbean. Harper, D.A.T. & Donovan, S.K. 1990. Fossil brachi- Geological Society of America Bulletin, 97, 48-60. opods of Jamaica Journal of the Geological Society of 30Draper, G. 1987. A revised tectonic model for the evolution Jamaica, 27, 27-32. of Jamaica: in Ahmad, R. (ed.), Proceedings of a 44Hendry, M.D. & Digerfeldt, G. 1989. Palaeogeography Workshop on the Status of Jamaican Geology. Geological and palaeoenvironments of a tropical coastal wetland Society of Jamaica, Special Issue, 120-150. and offshore shelf during Holocene submergence, Ja- 31Draper, G. 1990. Jamaica. (In Draper, G. & Lewis, J.F. maica. Palaeogeography, Palaeoclimatology, Pa- Geology and tectonic evolution of the northern Caribbean laeoecology,73,1-10. margin): in Dengo, G. & Case, J.E. (eds), The 45Hill, R.T. 1899. The geology and physical geography of Geology of North America, Volume H, The Caribbean Jamaica: a study of a type of Antillean development. Region, 120-127. Geological Society of America, Bulletins of the Museum of Comparative Zoology, Har- Boulder, Colorado. vard, 34, 1-256. 32Draper, G., Harding, R.R., Horsfield, W.T., Kemp, A.W. 46Hill, V.G. 1973. The bauxites of Jamaica: in Hill, V.G. & Tresham, A.E. 1976. Low -grade metamorphic belt (ed.), Proceedings of Bauxite Symposium II. Journal of in Jamaica and its tectonic implications. Geological the Geological Society of Jamaica, Special Issue, 1-5. Society of America Bulletin, 87, 1283-1290. 47Holliday, D.W. 1971. Origin of Lower Eocene gypsum- 33Eva, A.N. 1980. Petroleum potential of Jamaica: a case anhydrite rocks, southeast St. Andrew, Jamaica. Trans- study of part of an ancient island arc. UN ESCAP, actions of the Institute of Mining and Metallurgy,80, CCOP/SOPAC Technical Bulletin, 3, 143-151. B305-B315. 34Eva, A.N. & McFarlane, N.A. 1985. Tertiary to early 48Horsfield, W.T. 1972. A late Pleistocene sealevel notch, Quaternary carbonate facies relationships in Jamaica: and its relation to block faulting on the north coast of in Transactions of the 4th Latin American Geological Jamaica. Journal of the Geological Society of Jamaica, Conference, Port-of-Spain, Trinidad, 7th-15th July, 12, 18-22. 7979,1, 210-219. 49Horsfield, W.T. 1974. Major faults in Jamaica Journal of 35Gordon, C.M. & Donovan, S.K. 1992. Disarticulated echi- the Geological Society of Jamaica, 14, 1-4. noid ossicles in paleoecology and taphonomy: the last 50Horsfield, W.T. 1975. Quaternary vertical movements in interglacial Falmouth Formation of Jamaica. Palaios, the Greater Antilles. Geological Society of America 7,157-166. Bulletin, 86, 933-988. 36Goreau, T.F. & Burke, K. 1966. Pleistocene and Holocene 51Horsfield, W.T. & Robinson, E. 1976. Marine geology of geology of the island shelf near Kingston, Jamaica. the Jamaica Passage: in Causse, R. (ed.), Transactions Marine Geology, 4, 207-225. of the 7th Caribbean Geological Conference, Guade- 37Gose, W. A. 1985. Caribbean tectonics from a paleomag- loupe, 30th June-I2th July, 1974,107-113. netic perspective: in Stehli, F.G. & Webb, S.D. (eds), 52Horsfield, W.T. & Roobol, M.J. 1974. A tectonic model for The Great American Interchange, 285-301. Plenum the evolution of Jamaica. Journal of the Geological Press, New York. Society of Jamaica, 14, 31-38. 38Gose, W.A. & Testamarta, M.M.1983. Paleomagnetic 53Hose, H.R & Versey, H.R. 1956. Palaeontological and results from sedimentary rocks in Jamaica: initial results. lithological divisions of the lower Tertiary limestones of Journal of the Geological Society of Jamaica, 22, Jamaica. Colonial Geology and Mineral Resources; 16-24. 6,19-39. 39Green, G.W. 1977. Structure and stratigraphy of the Wag-water 54Isaacs, M.C. & Jackson, T.A. 1987. The mineralogy and Belt, Kingston, Jamaica. Overseas Geology and geochemistry of plutonic rocks from Jamaica: in Ah- Mineral Resources, 48, 1-21. mad, R. (ed.), Proceedings of a Workshop on the Status of 40Grippi, J. 1980. Geology of the Lucea Inlier, western Jamaican Geology. Geological Society of Jamaica, Jamaica Journal of the Geological Society of Jamaica, Special Issue, 95-106. 19, 1-24. 55Jackson, T.A. 1987. The petrology of Jamaican Creta- 41 Grippi, J. & Burke, K. 1980. Submarine canyon complex ceous and Tertiary volcanic rocks and their tectonic among Cretaceous island arc sediments, western Ja- significance: in Ahmad, R. (ed.), Proceedings of a maica. Geological Society of America Bulletin, 91, Workshop on the Status of Jamaican Geology. Geological 179-184. Society of Jamaica, Special Issue, 69-94. 42Harland, W.B., Smith, A.G. & Wilcock, B. (eds). 1964. 56Jackson, T.A. & Smith, T.E. 1979. Tectonic significance

124 EDWARD ROBINSON

71 of basalts and dacites in the Wagwater Belt, Jamaica Mann, P., Draper, G. & Burke, K. 1985. Neotectonics and Geological Magazine, 116, 365-374. sedimentation at a strike-slip restraining bend system, 57Jackson, T.A., Smith, T.E. & Isaacs, M.C 1989. The Jamaica: in Riddle, K. (ed.), Strike-slip Deformation, significance of geochemical variations in the Creta- Basin Formation, and Sedimentation. Society of Eco- ceous volcanic and plutonic rocks of intermediate and nomic Paleontologists and Mineralogists Special Pub- felsic composition from Jamaica. Journal of the Geo- lication, 37,211-226. logical Society of Jamaica, 26, 33-42. 72Matley, C.A. 1925. Recent geological work in Jamaica. 58Jiang, M.-J. & Robinson, E. 1987. Calcareous nannofos- Report to the British Association (Toronto), 1924, 391- sils and larger foraminifera in Jamaican rocks of Creta- 392. ceous to early Eocene age: in Ahmad, R. (ed.), 73Matley, C.A. 1951. The Geology and Physiography of the Proceedings of a Workshop on the Status of Jamaican Kingston district, Jamaica (ed. F. Raw). Institute of Geology. Geological Society of Jamaica, Special Issue, Jamaica; published by the Crown Agents, London, 139 24-51. pp. 59Kashfi, M.S. 1983. Geology and hydrocarbon prospects 74McFarlane, N.A. (compiler). 1977. Geological map of of Jamaica. American Association of Petroleum Geolo- Jamaica, 1:250,000. Ministry of Mining and Natural gists Bulletin, 67, 2117-2124. Resources, Kingston. 60Kauffinan, E.G. 1966. Notes on Cretaceous Inoceramidae 75Meloche, J.D. 1987. Regional thermal alteration of Creta- (Bivalvia) of Jamaica. Journal of the Geological Soci- ceous strata, Jamaica: fact or fiction?: in Ahmad, R. ety of Jamaica, 8, 37-40. (ed.), Proceedings of a Workshop on the Status of 61Kaufmann, E.G. & Sohl, N.F. 1974. Structure and evolu- Jamaican Geology. Geological Society of Jamaica, tion of Antillean Cretaceous rudist frameworks. Ver- Special Issue, 69-94. handlungen Naturforschenden Gesellschaft in Basel, 76Meyerhoff, A. A. & Krie g, E.A. 1977. Petroleum potential of 84, 399-467. Jamaica. Special Report for the Ministry of Mining and 62Kemp, A. W. 1971. The geology of the south-western flank Natural Resources, Kingston, 131 pp. of the Blue Mountains, Jamaica. Unpublished Ph.D. 77Pickerill, RK. & Donovan, S.K. 1991. Observations on thesis, University of the West Indies, Mona. the ichnology of the Richmond Formation of eastern 63 Kinghom, M. 1977. Bibliography of Jamaican Geology. Jamaica. Journal of the Geological Society of Jamaica, Geo Abstracts, Norwich, 150pp. 28, 19-35. 64 Krijnen, J.P. & Lee Chin, A.C. 1978. Geology of the 78Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of northern central and southestera Blue Mountains, Ja- the Caribbean region; a plate tectonic perspective: in maica, with a provisional compilation of the entire Dengo, G. & Case, J.E. (eds), The Geology of North inlier. Geologie en Mijnbouw, 57, 243-250. America, Volume H, The Caribbean Region. Geologi- 65 Land, L.S. 1991. Some aspects of the late Cenozoic evo- cal Society of America, Boulder, Colorado. lution of north Jamaica as revealed by strontium isotope 79Porter, A.R.D. 1970. Geology of the Ginger Ridge grano- study. Journal of the Geological Society of Jamaica, diorite stock and associated rocks. Unpublished M.Sc. 28, 45-48. thesis, University of the West Indies, Mona. 66 Larson, D.C. 1983. Depositional facies and 80Porter, A.R.D. 1990. Jamaica: a Geological Portrait. diagenetic fabrics in the late Pleistocene Falmouth Institute of Jamaica, Kingston, 152 pp. Formation of Jamaica. Unpublished M.S. thesis, 81Porter, A.R.D., Jackson, T.A. & Robinson, E. 1982. Min- University of Oklahoma, Norman. erals and Rocks of Jamaica. Jamaica Publishing House, 67 Lewis, J.F., Harper, C.T., Kemp, A.W. & Stipp, J.J. Kingston, 174pp. 1973. Potassium-argon retention ages of some 82Poulton, D. 1987. Petroleum source rock potential in Cretaceous rocks from Jamaica. Geological Jamaica: in Ahmad, R. (ed.), Proceedings of a Work- Society of America Bulletin, 84, 335-340. shop on the Status of Jamaican Geology. Geological 68 Lyew -Ayee, A. (ed.). 1982. Proceedings of Bauxite Society of Jamaica, Special Issue, 310-330. Symposium V. Journal of the Geological Society of 83Reed, A.J. 1966. Geology of the Bog Walk Quadrangle, Jamaica, Special Issue, 7, 328 pp. Jamaica. Geological Survey Department, Jamaica. 69 Mann, P. & Burke, K. 1984. Cenozoic rift formation Bulletin, 6, 54 pp. in the northern Caribbean. Geology, 12, 732-736. 84Robinson, E. 1960. Observations on the elevated and 70 Mann, P. & Burke, K. 1990. Transverse intra-arc modern reef formations of the St. Ann coast. Geonotes. rifting: Paleogene Wagwater Belt, Jamaica. Marine 3,18-22. and Petroleum Geology, 7, 410-427. 85Robinson, E. 1969a. Stratigraphy and age of the Dump

125 Jamaica

limestone lenticle, central Jamaica. Eclogae Geologi- cene), Jamaica, West Indies. Paleogeography, cae Hehetiae, 62, 737-743. Palaeo-climatology, Paleoecology, 16,217-242. 100 86Robinson, K 1969b. Geological field guide to Neogene Steineck, P.L. 1981. Upper Eocene to Middle Miocene sections in Jamaica West Indies. Journal of the Geo- ostracode faunas and paleo-oceanography of the logical Society of Jamaica, 10, 1-24. north coastal belt, Jamaica, West Indies. Marine 87Robinson, E. 1971. Observations on the geology of Micropa-leontology, 6,339-366. 101 Jamaican bauxite; in Bauxite/Alumina Symposium. Sweeting, M.M. 1958. The karstlands of Jamaica. Geo- Journal of the Geological Society of Jamaica, graphical Journal, 124,184-199. 102 Special Issue, 3-9. Trechmann, C.T. 1927. The Cretaceous shales of Ja- 88Robinson, E. 1976. Lignite in Jamaica. Ministry of Min- maica. Geological Magazine, 64, 27-65. 103 ing and Natural Resources, Kingston, 63 pp. Trechmann, C.T. 1930. The Manchioneal Beds of Ja- 89Robinson, E. 1988. Late Cretaceous and early Tertiary maica. Geological Magazine, 67,199-218. 104 sedimentary rocks of the Central Inlier, Jamaica. Jour- Versey, H.R. 1957. The White Limestone of Jamaica and nal of the Geological Society of Jamaica, 24, 49-67. thepalaeogeography governing its deposition. 90Robinson, E. 1993. Some imperforate foraminifera from Unpublished M.Sc. thesis, University of Leeds. 105 the Paleogene of Jamaica and the Nicaragua Rise. Jour- Vila, J.-M., Amilcar, H., Amilcar, H.C., Boisson, D. & nal ofForaminiferal Research, 23, 47-65. Feinberg, H., 1990. Un evenement tectono-sedimen- 91Robinson, E. & Cambray, F.W. 1971. Physiography of the taire Danian dans le sud d'Haiti (riviere Gosseline, sea floor south east of Jamaica: in Symposium on Inves- Grandes Antilles): consequences sur 1'extension et la tigations and Resources of the Caribbean Sea and mise en place de la nappe de Macaya. Bulletin de la Adjacent Regions, Curacao, 1968. UNESCO-FAO, Societe Geologique de France, series 8, 6, 349-359. 106 Paris, 285-289. Wadge, G. 1982. A Miocene submarine volcano at Low 92Robinson, E. & Jiang, M.M.-J. 1990. Paleogene calcare- Layton, Jamaica. Geological Magazine, 119,193-199. 107 ous nannofossils from western Portland, and the ages Wadge, G. & Dixon, T. 1984. A geologic interpretation and significance of the Richmond and Mooretown For- of SEASAT-SAR imagery of Jamaica. Journal of Ge- mations of Jamaica. Journal of the Geological Society ology, 92, 561-581. 108 of Jamaica, 27,17-24. Wadge, G., Draper, G. & Robinson, E. 1982. Gravity 93Robinson, E. & Lewis, J.F. 1970. Field guide to aspects anomalies in the Blue Mountains, eastern Jamaica: in of the geology of Jamaica: in Donnelly, T.W. (ed.), Snow, W., Gil, N., Llinas, R, Rodriguez-Torres, R., International Field Institute Guidebook to the Carib- Seaward, M. & Tavares, I. (eds), Transactions of the bean Island Arc System. American Geological Institute, 9th Caribbean Geological Conference, Santo Dom- Washington, D.C., 48 pp. ingo, Dominican Republic, 16th-20th August, 1980, 94Roobol, MJ. 1972. The volcanic geology of Jamaic a: 467-474. 109 in Petzall, C. (ed.), Transactions of the 6th Wadge, G. & Eva, A.N. 1978. The geology and tectonic Caribbean Geological Conference, Margarita, significance of the Sunning Hill Inlier. Journal of the Venezuela, 6th-14th July, 1971, 100-107. Geological Society of Jamaica, 17, 1-15. 110 95Roobol, M.J. 1976. Post-eruptive mechanical sorting of Wadge, G., Jackson, T.A., Isaacs, M.C. & Smith, T.E. pyroclastic material—an example from Jamaica. Geo- 1982.Ophiolitic Bath-Dunrobin Formation, Jamaica; logical Magazine, 113, 429-440. significance for the Cretaceous plate margin evolution 96Roobol, M.J. & Horsfield, W.T. 1976. Sea floor lava in the northwestern Caribbean. Journal of the Geologi- outcrop in the Jamaica Passage. Journal of the Geologi- cal Society of London, 139, 321-333. 111 cal Society of Jamaica, 15, 7-10. Westcott, W. & Etheridge, F.G. 1983. Eocene fan delta- 97Schmidt, W. 1988. Stratigraphy and depositional environ- submarine fan deposition in the Wagwater trough, east- ment of the Lucea Inlier, western Jamaica. Journal central Jamaica. Sedimentology, 30, 235-247. 112 of the Geological Society of Jamaica, 24, 15-35. Whittaker, A. and 11 others. 1991. A guide to strati- 98Scott, W. 1987. Age and provenance of the Richmond graphic procedure. Journal of the Geological Society Formation of the Rio Grande valley, eastern Jamaica: of London, 148, 813-824. in Ahmad, R (ed.), Proceedings of a Workshop on the 113Wright, R.M. 1971. Tertiary biostratigraphy of central Status of Jamaican Geology. Geological Society of Jamaica: tectonic and environmental implications: in Jamaica, Special Issue, 69-94. Mattson, P. (ed.), Transactions of the 5th Caribbean 99Steineck, P.L. 1974. Foraminiferal paleoecology of the Geological Conference, St. Thomas, Virgin Islands, Montpelier and lower Coastal Groups (Eocene - Mio- lst-5th July, 1968,129.

126 EDWARD ROBINSON

116 114Wright, R.M. (ed.). 1974. Field Guide to Selected Zans, V.A. 1952. Bauxite resources of Jamaica and their Jamaican Geological Localities. Ministry of development. Colonial Geology and Mineral Re- sources, 3,307-333 Mining and Natural Resources, Mines and Geology 117 Division, Kingston, Special Publication, 1, 57 pp. Zans, V.A., Chubb, L.J., Versey, H.R., Williams, J.B., Robinson, E. & Cooke, D.L. 1963. Synopsis of the 115Wright, R.M. & Robinson, E. (eds). 1993. geology of Jamaica. Geological Survey of Jamaica Biostratigra-phy of Jamaica. Geological Society of Bulletin, 4, 72 pp. America Memoir, 182, 492 pp.

127

CHAPTER 7

Hispaniola

GRENVILLE DRAPER1, PAUL MANN2 and JOHN F. LEWIS3

1 Department of Geology, Florida International University, Miami, Florida 33199, U.S.A. 2Institute for Geophysics, University of Texas at Austin, Austin, Texas 78759, U.S.A. 3Department of Geology, George Washington University, Washington, D. C., 20052, U.S.A.

INTRODUCTION which forms the basement for late Tertiary sedimentary basins. The overall geology is illustrated in Fig. 7.2. HISPANIOLA, WHICH is divided politically between the The Cretaceous-early Tertiary rocks, whose outcrop countries of the Dominican Republic and Haiti, has an area covers about 30% of the area of the island, formed as aresult of about 80,000 square km, making it the second largest of the interactions on the plate boundary that separated the island in the Greater Antilles. Physiographically, the island proto-Caribbean and North American Plates. Two major is composed of a four rugged, westnorthwest-eastsoutheast groups of rocks occur. trending mountain ranges separated by three lower lying valleys (Fig. 7.1). The twin peaks of Pico Duarte and La (1) In the northern part of the island there is an early Pelona (3087 m) form the highest elevation in the Greater Cretaceous-Eocene island arc assemblage of rocks. Antilles. The steep topography of the island hints as to its This assemblage, or collage, has components of the recent tectonic (neotectonic) activity and, indeed, the island fore-arc, magmatic arc, the oceanic basement of the arc, has many recently active faults and uplifted areas, and is a possible closed back arc basin of late Cretaceous age seismically active (Fig. 7.264). Major earthquakes have been and a later Cretaceous-Eocene back-arc basin and rem experienced in 1751, 1770, 1842, 1887, 1911, 1946, 1948 nant arc. and 195375. Besides being a critical area for studies of the neotec- (2) A late Cretaceous oceanic plateau forms the basement tonics (Miocene and younger crustal movements) of the of the southwestern part of the island and is the uplifted northern Caribbean plate, Hispaniola also records part of the edge of crust that underlies much of the Caribbean Sea Cretaceous-early Tertiary island arc and late Tertiary strike- 18 slip history of the region . Despite its geological impor- Rocks of late Eocene and younger age consist mainly tance, the details of the geology of many regions in of clastic and carbonate rocks deposited in sedimentary Hispaniola were known only sketchily before the 1980s. basins which formed as a result of extensive east-west, left Numerous studies during the 1980s have increased our lateral strike-slip tectonics that have affected and continue knowledge of the geology and tectonics of the island. Much to affect the island. Pliocene to Pleistocene alkaline volcanic of the new work in the Dominican part of the island is rocks that occur in the Cordillera Central are also related to 66 summarized in arecent volume of collected papers . Other strike-slip tectonic activity. summaries of the geology of Hispaniola can found in Bowin10, Lewis56, Lewis and Draper57, Lidz and Nagle61, and 70 TERRANES AND MORPHOTECTONIC ZONES Maurrasse . IN HISPANIOLA

Definition of terranes OUTLINE OF GEOLOGY A characteristic of the geology of Hispaniola is that it Hispaniola comprises a Cretaceous -early Eocene substrate is made up of several fault-bounded blocks, or geological

129 Hispaniola

Figure 7.1. Map showing physiographic provinces and locations in Hispaniola (modified from Weyl103, Lewis and Draper57). provinces, each of which has a distinct geological history origin of each terrane as either oceanic, fore-arc, magmatic that differs from that of adjacent blocks. How the geological arc or back arc. Several of the terrane boundaries are either evolution of one block relates to that of an adjacent block is completely or partially covered by thick, clastic sedimentary not always intuitively obvious. In the North American Cor- basins formed during a late Miocene to Recent transpres- dillera, which has similar characteristics, these provinces are sion. In Table 7.2, we speculate on the identity of the known as tectonostratigraphic terranes. Therefore, we fol- terranes underlying these transpressional basins. low Mann et al.67 in describing the components of His- paniola's geology in terms of 12 distinct terranes (Fig. 7.3a; Morphotectonic zones of Hispaniola Table 7.1) Earlier descriptions of the geology of Hispaniola56,57 We define Hispaniola's terranes using the criteria of divided the island into 10 morphotectonic zones (Fig. 7.3b; Howell et al.49: "A tectonostratigraphic terrane is a fault- Table 7.2). This division preceded the application of the bounded package of rocks of regional extent characterized terrane concept and should not be confused with it. The by a geologic history which differs from that of neighboring morphotectonic zones correspond to the physiographic terranes. Terranes may be characterized internally by a provinces of the island and, therefore, are the fault bounded distinctive stratigraphy, but in some cases ametamorphic or blocks and basins produced by Neogene tectonic activity. tectonic overprint is the most distinctive characteristic.... In The strong, but not complete, correspondence of terranes general, the basic characteristic of terranes is that the present with morphotectonic zones results from reactivation of pre- spatial relations are not compatible with the inferred geo- existing terrane boundaries66. logic histories... Terranes can be grouped into three catego- ries: stratigraphic, disrupted, and metamorphic." Northern Hispaniola: Samand metamorphic terrane It is important to point out that herein we define the (fore-arc) terranes of Hispaniola purely descriptively. There is no The terranes are described in the order encountered implication that the terranes we describe are neccessarily traversing the island from northeast to southwest (Figs. 7.2, far-travelled, or 'suspect', to use the jargon of the terranolo- 7.3a). The basement rocksof the Samanapeninsula consists gists. mainly of a coherent metamorphic terrane characterized by In the following description of individual terranes, it the presence of lawsonite and/or pumpellyite. Lithologi- can be seen that Hispaniola is comprised of stratigraphic , cally, most of the Cretaceous rocks of the rocks of Samana metamorphic and disrupted terranes. These terranes are consist of intercalated mica schists and marble bands50,51. In typically much longer than they are wide and are bounded the southeastern margin of the terrane, a 2 x 10 km region by high-angle faults, with strike-slip and/or reverse compo- (the PuntaBalandrazone) contains blueschist, eclogite50,79, nents. In addition, we have attempted to indicate the tectonic calc -silicate gneisses and calcite-siderite metasomatites.

130 G. DRAPER, P. MANN and J.F. LEWIS

Figure 7.2. Generalized geological map of Hispaniola (modified from Lewis and Draper 57).

131 Hispaniola

The last occur as lenses and layers intercalated with the mica a basaltic flow. schists and also as blocks in a small, melange-like body34. Metre-scale slivers of serpentinite also occur50,51. Joyce51 Northern Hispaniola: Altamira stratigraphic terrane estimated that prograde recrystallization of blueschist and (magmatic arc) eclogite assemblages took place at 400-500°C and 10 Kb. The oldest exposed rocks of this terrane consist of Sm-Nb and K-Ar whole rock ages from the eclogite in the biomicritic limestones interbedded with minor amounts of melange body suggest that the high pressure assemblages volcaniclastic rocks, tuffaceous siltstones and mudstones equilibrated in the late Cretaceous, but that the blocking containing an Upper Paleocene through Lower Eocene mi- temperature of the mica was not reached until late Eo- crofauna (Los Hidalgos Formation24). The biomicrites are cene51,86. intruded by porphyritic dykes and sills (Palma Picada For- Metamorphic rocks of the Samana terrane are uncon- mation78). An angular unconformity marked by an Upper formably overlain by Middle to Upper Miocene shallow- Eocene basal conglomerate truncates both the Los Hidalgos water limestones. and Palma Picada Formations. Both the Rio San Juan/Puerto Plata/Pedro Garcia and Northern Hispaniola: Rio San Juan/Puerto Plata/Pedro Altamira terranes are overlain by Upper Eocene to Lower Garcia disrupted terrane (fore-arc) Miocene marine conglomerates and turbididtic sandstones of This terrane consists of a heterogeneous assemblage of the Mamey Group. De Zoeten and Mann24 identified four igneous and metamorphic rocks which outcrop in three main separate formations in this group, each of which seems to inliers along the Atlantic coast of the island. It is likely that have been deposited in a separate basin. The famous amber more detailed studies will further sub-divide the heteroge- of the Dominican Republic are largely from turbiditic sand- neous rocks of these three widely separated inliers (Fig. 7.2). stones of the Mamey Group. Unconformably overlying the The Rio San Juan-Puerto Plata-Pedro Garcia terrane is Mamey Group are Upper Miocene marls and Upper Mio- faulted against the Altamira terrane (see below) along the cene to Pliocene reefal limestones (Villa Trina Formation). Rio Grande fault zone24. The eastern inlier, near the town of Rio San Juan, contains four major units: blueschist-eclo- Eastern Hispaniola: Seibo stratigraphic terrane gite melanges with serpentinite matrix; fine-grained cohrent (magmatic arc) greenschist-blueschist facies rocks; coarse-grained amphi- The Seibo terrane in the Cordillera Oriental of the bolite facies rocks; and a gabbroic intrusive complex34-36. eastern Dominican Republic includes some of the oldest Draper and Nagle35 suggested that the northern part of this reliably dated rocks in the island. These are the hydrother- area was the result of subduction zone metamorphism, while mally metamorphosed volcanic rocks of the Los Ranchos the southern part represents the basement of the Hispaniola Formation7,9,54 that include pillowed and non-pillowed basalts, forearc. The terrane was covered by Upper Eocene to Mio- dacites, keratophyres, rhyolites, andesites, tuffs and cene clastic marine sedimentary rocks following the cessation volcanic breccias. Chemically, they have compositions of subduction in the middle Eocene. characteristic of the primitive island arc (PIA) series27,28,55. The basal rocks of the western inlier of the terrane near The upper part of the Los Ranchos Formation appears to Puerto Plata consist of serpentinite, gabbro, and volcanic have been emergent in the final phases of volcanism91, rocks80,88. Large gabbroic and dioritic blocks are common possibly because they accumulated on a shallow substrate of inclusions in the serpentinite, which was intruded by dykes of previously thickened crust. rodingitized gabbro in other areas. Andesite flows, spilitic The Los Ranchos Formation is the host of rich gold pillow basalts (associated with Lower Cretaceous radiolari- mineralization. The mineable reserves are formed of resid- ans in sediment between pillows; H. Montgomery and E. ual (weathered) deposits of hydrotheimally altered black Pessagno, personal communication), hyaloclastites and shales of the lower Los Ranchos Formation. Until recently, tuffs of the Los Canos Formation are associated with the the Pueblo Viejo mine was the largest gold mine in the serpentinite-gabbro complex. Overlying these rocks is a 1 western hemisphere and the largest open-pit gold mine in km thick section of interbedded conglomerates, sandstones the world. and white, dacitic tuffs of the Imbert Formation. Some thin The basal volcanic and plutonic section of the eastern chert and limestone beds have yielded fossils which indicate a Seibo terrane is truncated by an erosional unconformity. The Paleocene-Early Eocene age. unconformity is overlain byplatform carbonates, containing a The basal rocks of the Pedro Garcia inlier consist of tuff rich Aptian- Albian fauna , which in turn are overlain by and mafic amygdaloidal lava which is intruded by basaltic Upper Cretaceous volcaniclastic marine sedimentary rocks, dykes and a small tonalite stock24,85. Bowin and Nagle11 tuffs, potassium-rich basalt flows and pelagic limestones reported a Maastrichtian whole rock K-Ar age of 72 Ma for containing Upper Cenomanian to Lower Turonian fossils55.

132 G. DRAPER, P. MANN and J.F. LEWIS

Figure 7.3. (A) Morphotectonic zones of Hispaniola (from Lewis ; Lewis and Draper57). Key to numbered zones: Zone 1 = Old Bahama Trench; Zone 2 = Cordillera Septentrional-Samana Peninsula; Zone 3 = Cibao Val- ley; Zone 4 = Massif du Nord-Cordillera Central; Zone 5 = Northwestern-south-central zone which includes the following features: Plateau Central-San Juan Valley-Azua Plain; Sierra el Numero; Presqu'ile de Nord-Ouest; Montaignes Noires; Chaines de Matheux-Sierra de Neiba; and Sierra de Martin Garcia; Zone 6 = He de la Go- nave-Plaine de Cul-de-Sac-Enriquillo Valley; Zone 7 = southern peninsula which includes Massif de la Selle- Massif de la Hotte-Sierra de Bahoruco; Zone 8 = eastern peninsula which includes Cordillera Oriental and the Seibo coastal plain; Zone 9 = San Pedro basin and north slope of the Muertos Trough; and Zone 10 = Beata Ridge and southern peninsula of Barahona. (B) Tectonic terranes of Hispaniola. See text and Table 7.1 for a correlation of morphotectonic zone and tectonic terranes.

Several whole rock K-Ar dates on basalts of the eastern taining ultramafic clasts7. Most of the rocks of the terrane Seibo terrane have yielded an age range from Aptian to possess a pervasive planar slaty cleavage. The limestone Santonian age1. The volcanic sequence of the eastern Seibo beds contain well-preserved ammonites of early Coniacian terrane is intruded by a large granodiorite pluton which has age and a chert bed has yielded a rich fauna of Coniacian yielded a Santonian whole rock K-Ar age. radiolarians8. Bourdon7 interpreted both the Oro and Seibo terranes Eastern Hipaniola: Oro stratigraphic terrane as sedimentary basins of approximately the same age, but of (magmatic arc) differing sedimentary facies, that formed above the mag- The Oro terrane consists of a folded 1500 m thick matic part of an island arc. According to Bourdon7, the section of volcaniclastic sedimentary rocks with minor Seibo terrane overthrust the Oro terrane during the middle black limestones, radiolarian cherts and conglomerates con- Eocene.

133 Hispaniola

Central Hispaniola—the Median Belt: Tortue-Amina- the Hispaniola fault zone9,25. Maimon metamorphic terrane (magnetic arc) The serpentinite is in contact with massive, unfoliated This terrane consists of a long (300 km), narrow (5-15 basalt (Peralvillo Formation) and basalt with a large fraction km) outcrop of metamorphic rocks extending diagonally of hyaloclastic material (Siete Cabezas Formation). The across central Hispaniola from northwest to southeast. latter formation also contains chert which has yielded 5,26 Three main areas of outcrop occur: Tortue Island off of the Cenomanian-Santonian age radiolarians . Because of north coast of Haiti81,101; the Amina area on the northern faulting, the stratigraphic relationship between the basaltic flank of the Cordillera Central of the Dominican Repub- and ultramafic lithologies is unknown. lic 31,33; and the Maimon area of the eastern Cordillera The Loma Car ibe serpentinite was interpreted by Lewis 59 Central31,33,53. and Jimenez to be serpentinized harzburgitic oceanic man- 39 On the Ile de la Tortue, the metamorphic rocks are tle, which was confirmed by geochemical studies . The well-foliated meta-tuffs and meta-rhyolites (quartz-albite- basaltic and ultramafic rocks can thus be interpreted as epidote-chlorite schists) intercalated with white marbles 81. forming part of a dismembered ophiolite complex. Until The calcareous schists, which contain a Cenomanian to recently, it appeared that the extensive gabbroic component Campaman microfauna101, are unconformably overlain by present in most ophiolite complexes was absent in the Loma Paleocene-Eocene limestones. In the Amina and Maimon Caribe terrane, but recent fieldwork by Draper and Lewis areas the terrane consists mainly of meta-igneous rocks of (unpublished data, summer 1991) indicates that the area east basaltic to quartz keratophyric composition and meta-sedi- of the serpentinite belt, mapped as 'Duarte Formation' by 9 mentary lithologies which include quartzite, metaconglom- Bowin , consists mainly of an assemblage of gabbroic rocks erate, graphitic schist, and marble. Draper and Lewis31,33 that we have informally named the Rio Verde complex (Fig. concluded that the rocks of this assemblage are metasedi- 7.2) mentary rocks of probable volcanogenic origin that may The northwestern part of the Loma Caribe/Tavera ter- have formed in the magmatic part of an island arc. In the rane, which runs along the northern flank of the Cordillera Amina area, the internal structure is characterised by tight Central, consists of folded and faulted Oligocene-Lower northeast-verging folds 31, and recent investigations in the Miocene conglomerates and turbidites of the Tavera Group Maimon area (Draper and Lewis, unpublished data, summer that outcrop in a narrow 60 km long by 5-15 km wide 25,44,84 1991) have revealed northeast-verging fold and thrust struc- sedimentary basin . Although the exposed contact with tures of probable Cretaceous age. the ophiolitic rocks is faulted, we interpret the Oligocene- The southern boundary of the terrane consists of high- Lower Miocene sedimentary basin as originally having angle faults, at least one of which is overlain by Upper been deposited above the ultramafic and basaltic Miocene clastic sedimentary rocks33,8 4. In the Maimon area basement rocks. The basin most probably developed as a the terrane is separated from the Seibo terrane by the south- pull-apart trough in response to Oligocene-early Miocene east dipping Hatillo thrust9, but it is not clear how far the left-lateral strike-slip movement25. Amina segment extends beneath Neogene sedimentary The southern contact of the Loma Caribe-Tavera ter- rocks of the Cibao Valley (Fig. 7.3B). rane with the Duarte terrane is a high-angle fault that is locally covered by the Oligocene Moncion Limestone84. Central Hispaniola—the Median Belt: Loma The northern contact with the Amina-Maimon-Tortue ter- Caribe/Tavera ophiolitic(?) terrane (oceanic) rane is also a high-angle fault which is onlapped by an Upper The Loma Caribe-Tavera terrane consists of a 150 km Miocene basal conglomerate. long by 5-15 km wide fault-bounded belt that diagonally crosses central Hispaniola from northwest to southeast and Central Hispaniola—the Median Belt: Duarte forms the centre line of the Median Belt of Bowin9. (The metamorphic terrane (oceanic) median belt also includes the flanking metamorphic belts of The Duarte terrane outcrops along the northern flank of the Duarte and Amina-Maimon-Tortue terrenes.) The ter - the Cordillera Central in the Dominican Republic and con- rane is formed of two Upper Cretaceous basalt units and a sists of massive, locally schistose meta-volcanic rocks of serpentinized harzburgite/dunite body that together outcrop mafic to ultramafic composition which are regionally meta- along the eastern flank of the Cordillera Central. morphosed to grades ranging from prehnite-pumpellyite to 32,33,59 The faults bounding the serpentinite dip inward, making amphibolite facies . 33 59 the body narrower at depth than at the surface. The Draper and Lewis and Lewis and Jimenez have serpentinite was probably initially exposed and weathered divided the Duarte terrane into two units: a lower unit of in the Miocene45,59. Uplift may have initiated in the Oligo- greenschist facies mafic, high magnesian lava flows and cene and was possibly related to strike-slip movement along sills; and an upper unit of thin flows, pillow basalts and

134 G. DRAPER, P. MANN and J.F. LEWIS

Figure 7.4. Map showing topography and major Neogene faults of Hispaniola (modified from Mann et al.65) and lines of regional cross sections shown in Figure 7.6. Numbers identify major faults and fault-related features: 1 = Camu Fault Zone; 2 = Maimon Graben; 3 = Rio Grande Fault Zone; 4 = Septentrional Fault Zone; 5 = San Francisco Ridge; 6 = South Samana Fault Zone; 7 = Yabon Fault Zone; 8 = Hatillo Thrust Fault; 9 = Bonao Fault Zone; 10 = San Jose- Restauracion Fault Zone; 11 = Banilejo Fault Zone; 12 = Los Pozos -San Juan Fault Zone; 13 = Sierra de Neiba Fault Zone; 14 = Yayas-Constanza Plio-Pleistocene basalt province; 15 = Beata Fault Zone; 16 = Muertos Trench; 17 = Enri-quillo-Plantain Garden Fault Zone; 18 = Port-au-Prince Graben. Wavy lines indicate zone of late Neogene subduction accretion along the Muertos Trench. ______

Figure 7.5. (next page) Regional cross sections (not balanced) across central and eastern Hispaniola (Dominican Repub- lic) showing the relation between morphotectonic zones and tectonic terranes. Cross sections have no vertical exaggeration and are based on geologic data plotted on 1:100,000 compilation maps included as enclosures in Mann et al66. Tadpole symbols indicate measured bedding dips in the line of the cross section. The interpretation of structure below the uppermost several km is speculative. Key to abbreviations of lithologic formations: K1c-d = Cretaceous La Cienaga- Dumisseau Formation; Tsr = Paleocene-Lower Eocene San Rafael Formation; Tn = Middle-Upper Eocene Neiba For- mation; Tj-1 = Lower-Middle Oligocene Jeremie-Lemba Formation; Ts = Lower to Middle Miocene Sombrerito Formation; Tt = Upper Miocene-Lowermost Pliocene Trinchera Formation; Tab = Lower-Middle Pliocene Arroyo Blanco Formation; Tim = Pleistocene Las Matas Formation; TV = Vallejuelo Formation; Tve = Lower -Middle Eocene Ventura Formation; Tj = Middle Eocene Jura Formation; Ten = Middle-Upper Eocene El Numero Formation; Kt = Santonian Tireo Formation; D = Duarte Complex; Am = Amina schist; Tc = Cercado Formation; Tg = Gurabo Forma- tion; Tm = Mao Formation; Qa = Quaternary alluvium; KTlh = Upper Cretaceous-Middle Eocene Los Hidalgos Forma- tion; Kca = Cretaceous Cuaba amphibolites; Krb = Cretaceous Rio Boba intrusive suite; Keg = Cretaceous El Guineal Schists; Kjc = Upper Cretaceous Jagua Clara melange; S = serpentinite; Tcb = Upper Eocene-Lower Oligocene Canada Bonita Member of the Altamira Formation; Kpg = Upper Cretaceous Pedro Garcia Formation; Kxb = Puerto Plata basement complex; Tsm = post-Middle Miocene San Marcos Formation; Klg = Upper Cretaceous Las Guajabas Formation; Klgs = Upper Cretaceous siliceous horizons of the Las Guajabas Formation; Kc = Upper Cretaceous Rio Chavon For- mation; Kb = Upper Cretaceous Bejucalito Formation; Krc = Upper Cretaceous Rio Cuaron Formation. Key to fault abbreviations: BAFZ = Baharona fault zone; EPGFZ = Enriquillo-Plantain Garden fault zone; SJLPFZ = San Juan-Los Pozos fault zone; BFZ = Bonao fault zone; HFZ = Hispaniola fault zone; SFZ = Septentrional fault zone; GFZ = Guacara fault zone; RGFZ = Rio Grande fault zone; CFZ = Camu fault zone; San Jos e-Restauracion fault zone.

135

136 G. DRAPER, P. MANN and J.F. LEWIS

intrusive dolerites intercalated with tuffaceous and sedi- trichtian marine sedimentary rocks of the Trois Rivieres- mentary rocks of prehnite-pumpellyite facies. The foliation in Peralta Group (see below). As with the Duarte terrane, the Duarte terrane, although not always well-developed, is several large, unfoliated granitoid stocks and batholiths 53. generally steeply inclined and parallel to the outcrop of intruded the Tireo terrane up to the late Eocene theterrane. Based on the association of mafic lavas, pillow lavas, Central Hispaniola: Trois Rivieres-Peralta stratigraphic hyaloclastites and pelagic sedimentary rocks, Bowin10 and terrane (back arc basin) Palmer84 suggested that the Duarte terrane may be a fragment This terrane, which extends 320 km from northern Haiti to of metamorphosed oceanic floor. However, trace element the south-central Dominican Republic, consists of pack- analyses indicate that the Duarte terrane formed as part of a ets of marine turbiditic sandstones, siltstones and lime- seamount or ocean island32,59 that was later intruded and stones, ranging in age from Coniacian to Pleistocene. The overlain by late Cretaceous-Eocene island arc magmatic rocks. terrane is separated to the southeast from the Presqu'ile du Radiolarian fossils recovered from meta-cherts in the upper Nord-Ouest terrane by the San Juan-Los Pozos fault zone Duarte sequence indicate a late Jurassic age for the and to the northwest from the Tireo terrane by the San deposition of the Duarte Complex (H. Montgomery and E. Jose-Restauracion fault zone25. The belt can be divided into Pessagno, pers. comm., 1992). four distictive stratigraphic/structural packages. The several granitoid batholiths and stocks that intrude The oldest rocks in the terrane are highly faulted and the Duarte terrane were emplaced between the Albian and folded slivers of Coniacian through Danian micritic lime- late Eocene21,54. stones, tuffs, cherts, turbiditic sandstones and siltstones, and cherts of the Trois Rivieres Formation, which are found in Central Hispaniola: Tireo stratigraphic terrane scattered outcrops in northwestern Haiti6. These rocks ex- (magmatic arc) tend into the Dominican Republic, where similar Cam- The Tireo terrane forms a 290 km belt of Upper Creta- panian to low Upper Eocene lithologies have been ceous volcanic and epiclastic rocks which underlie the central mapped25,40,58,60. Some of the rocks have been metamor- mountain chain of Hispaniola and forms a roughly phosed to slates and are deformed into chevron folds with triangular area bounded on the north by the the Bonao-Gua- axial planar cleavage (Draper, unpublished observations). cara fault zone and on the south by the San Jose-Re- Detailed studies by Dolan et al.25 have subdivided the stauracic n fault zone. stratigraphy of the southeastern end of the the Trois Lewis et al.58 divided the rocks of the Tireo terrane into Rivieres-Peralta terrane into three packages consisting of: two units. The lower Tireo Group in the Massif du Nord of (1) the Paleocene-low Upper Eocene Peralta Group; (2) the Haiti and the Cordillera Central of the Dominican Republic Middle Eocene to Lower Miocene Rio Ocoa Group; and (3) consists of thick (about 4000 m) of mafic tuffs and high TiO2 the Lower Miocene to Pleistocene Ingenio Caei Group. Each basalts intercalated with mudstones, siltstones, cherts and group is separated from more deformed, underlying group limestones. Microfossil age determinations from the lower by major angular unconformities or faults46,47. Witschard Tireo Group are imprecise, but indicate an age in the range and Dolan104 and Dolan et al.25 interpreted the terrane as a Turonian to ?Campanian9. Rocks of the lower Tireo Group back-arc basin modified into an accretionary prism during correlate with andesites, tuffs, agglomerates, mudstones and oblique Eocene northeastward underthrusting of southwest- basalts of the Terrier Rouge series of the Massif du Nord of ern Hispaniola beneath central Hispaniola. Haiti82,83. The upper Tireo Group consist of an unknown thickness of lavas, volcanic breccias and tuffs of dacitic, Southwestern Hispaniola: Presqu'ile de Nord-Ouest Neiba rhyolitic and keratophyre composition. Felsic volcanic terrane (magmatic arc) plugs occur along the southern margin of the Tireo terrane. In the northwestern part of this terrane (Presqu'ile du Microfossil age determinations from intercalated sediments Nord-Ouest of Haiti), small inliers in the Tertiary cover indicate that the upper Tireo was probably deposited in the expose Upper Cretaceous intermediate volcanic rocks in- late Santonian to early-mid Campanian. 40Ar/39Ar dating has truded by small, Upper Cretaceous to Lower Eocene plu- yielded a Campanian age for the volcanic rocks in the tons 19,52. The Cretaceous rocks are unconformably overlain upper Tireo Group. In the Massif du Nord of Haiti, rocks of by Middle Eocene limestones with localized intercalations the upper Tireo Group correlate with dacite flows and of basaltic lava flows. These are in turn overlain by Oligo- stocks, crystal-lithic tuffs, and pyroclastic and volcaniclas -tic cene-Miocene clastic sediments of the Crete Formation. rocks of the La Mine Series 82,83. Cretaceous igneous rocks are not apparently exposed in In the south-central part of the terrane, the upper Tireo the regions to the southwest of the Presqu'ile du Nord (that Group pass conformably into Middle Campanian-Maas- is, in the Plateau Central, Montagnes Noires, Chaines des

137 Hispaniola

Figure 7.6. Tectonic model illustrating possible Campanian choking of a northward dipping subduction zone. Closure of a back arc basin to form the Median Belt ophiolite and reversal of subduction polarity and the reversal of polarity of subduction are a consequence of this plateau-arc collision (from Draper and Lewis33). Key to abbreviations: T=Tireo Group; D=Duarte Complex, SC=Siete Cabezas basalts; AM=Amina-Maimon Schists, LR=Los Ranches Formation; RSJ=Rio San Juan Complex; t=tonalite and other granitoid intrusions.

138 G. DRAPER, P. MANN and J.F. LEWIS

Matheux and Sierra de Neiba), although some Upper Creta- 1500 m thick section of interbedded pillowed and non-pil- ceous pelagic limestones occur in the Montagnes Noires 100. lowed basalts, dolerites, pelagic limestones, cherts, sili- Most of the Montagnes Noires-Chaine de Matheux-Sierra ceous siltstones and volcanogenic sandstones (Dumisseau 73 de Neiba consists of Eocene limestones and argillites inter - Formation ). Foraminifera and radiolaria recovered from calated with andesitic breccias, tuffs and rare basalts. The sedimentary interbeds yield ages from the late early Creta- Eocene rocks are overlain by Oligocene chalky and siliceous ceous to Turonian. K/Ar whole rock age determinations on limestones of the Madame Joie-Sombrerito Formation. samples of basalt and dolerite range in age from Albian to 1,93,96 The northern part of the terrane is occupied by the Maastrichtian . The seismic reflector B” underlying Plateau Central-San Juan Valley which contains more than much of the Caribbean Sea is composed of tholeiitic basalts 7600 m of Lower Eocene to Pliocene sedimentary rocks. In and dolerite sills intercalated with and overlain by pelagic the extreme southeast of the belt, these rocks are folded and sedimentary rocks of Coniacian and Campanian age . Be- thrust in a southwesterly direction. Biju-Duval et al.2 have cause of the geochemical similarity of these rocks with the 73 suggested that these structures form the leading edge of an basalts of the Dumisseau Formation, Maurrasse et al. and 94 accretionary prism structure which extends offshore into the Sen et al. suggested that is an uplifted piece of the Carib- San Pedro basin and Muertos Trough, and was formed by bean sea floor. Trace element and isotopic data from the the underthrusting of the Caribbean Sea floor under south- Massif de la Selle basalts further suggest a combined spread- 94 eastern Hispaniola. ing ridge and hotspot origin for the magmatism . The oldest rocks of the Sierra de Bahoruco of the Southwestern Hispaniola: Hotte-Selle-Bahoruco Dominic an Republic consist of small, highly faulted out- stratigraphic terrane (oceanic—Uplifted Caribbean crops of pillowed and non-pillowed basalts which have sea floor, B") yielded Maastrichtian K-Ar whole rock ages 96. The geo- The oldest (basement) rocks of this 350 km long terrane chemistry of these rocks indicates that, like those in Haiti, consist of Upper Cretaceous basalts intercalated with pe- they are non-orogenic tholeiites similar to those drilled at lagic sedimentary rocks which have been correlated with the Caribbean DSDP sites 42. rocks of the Caribbean Sea floor73. From west to east, the Late Cretaceous and Cretaceous/Tertiary boundary largest exposures of these rocks occur in the cores of three sequence breached en echelon anticlines that correspond to major In some parts of the Massif de la Selle an erosional mountain ranges on the southern peninsula of Hispaniola unconformity truncates the Dumisseau Formation22,99 and From west to east, these ranges axe the Massif de la Hotte marks the end of basaltic volcanism in the area The rocks and Massif de la Selle in Haiti, and the Sierra de Bahoruco above the unconformity surface are composed of Lower to in the Dominican Republic (Fig. 7.2). The Hotte-Selle-Ba- Middle Maastrichtian conglomerates and sandstones. How- horuco terrane is separated from the Presqu'ile du Nord- ever, in other areas of the Massif de la Selle, carbonate Ouest terrane to the north by the left-lateral sedimentation spans the Cretaceous-Tertiary boundary71. Enriquillo-Plantain Garden fault zone. The southern bound- Like other Cretaceous -Tertiary boundary sites this one has ary of the terrane is formed by fault zones offshore of the the high indium anomaly which has been attributed to a southern margin of the peninsula43. large meteorite impact. Moreover, it has been suggested that an The oldest rocks of the Massif de la Hotte are mainly arenaceous layer at the boundary in the Massif de la Selle, highly folded and fractured pelagic limestones and cherts which contains abundant shocked quartz, may be an impact ranging from Coniacian to Maastrichtian (Macaya Forma- surge deposit and thus may indicate that the impact site may tion16). These rocks are similar in lithology and age to have been in the Caribbean4,48,72,74. pelagic limestones recovered from deep-sea drilling sites in Pelagic limestone and chert of Albian to Maastrichtian the Colombian and Venezuelan basins38. The limestone age also outcrops in the Sierra de Bahoruco (Rio Arriba section is faulted against outcrops of pillowed and non-pil- Formation62), but the stratigraphic relation of this sedimentary lowed basalt whic h have yielded whole-rock K-Ar ages section to the underlying basalt is unknown. from Maastrichtian to Paleocene16. The Macaya Formation Tertiary succession and these basalts are overlain unconformably by a Lower The Paleocene-Miocene section of the Hotte-Selle-Ba- Paleocene conglomerate (Riviere Glace and Baraderes For- haruco terrane is characterized by gradually deepening car- mations;. In general this unconformity marks the end of bonate environments . Paleocene-Eocene carbonate rocks of basaltic volcanism in the Massif de la Hotte, although scat- the Siena Martin Garcia are similar to those in the Sierra de tered outcrops of alkalic volcanic rocks are interbedded in Bahoruco and are included as part of the Hotte-Selle-Ba- Upper Paleocene to Middle Eocene sedimentary rocks17. horuco terrane20 (Fig. 7.2). Rapid facies changes in carbon- In the Massif de la Selle, the oldest rocks consist of a ate environments during middle Miocene time are

139 Hispaniola

Figure 7.7. Terrane correlation chart and summary of tectonic events in Hispaniola. See text for explanation.

73 3 interpreted by Maurrasse et al. and Bizon et al as dating lated Cretaceous deformational events are not clear on the the initiation of uplift of the terrane and accompanying block cross sections as they are obscured by Cenozoic deforma- and/or strike-slip faulting. By the Pliocene, most of the tion. terrane was emergent with continental deposition in small, 68 The cross sections illustrate several important features strike-slip fault-controlled basins . about the structural history of the island:

REGIONAL STRUCTURE (1) The most prominent folding and thrusting event in south-central Hispaniola was late Miocene and younger The present-day structure of Hispaniola records the super - in age, and verges southward. This south vergence is position of various Cretaceous to Recent tectonic events reflected in the asymmetric topographic profile of the affecting its various terranes. In order to illustrate the overall Massif du Nord-Cordillera Central mountain range structure in Hispaniola, we present a series of four cross which has steeper slopes along its southwestern flank sections through the central and eastern parts of the island (Fig. 7.5). (Figs 7.4,7.5). In northeastern and central Hispaniola, the most promi- (2) The three major valleys of Hispaniola are either ‘fore- nent structural effects seen in the cross sections are due to land' basins, due to oblique thrusting on one side of the the middle to late Eocene collision between Hispaniola and basin (Cibao valley), or are 'ramp' basins caused by late the Bahama Platform. In southwestern Hispaniola, however, Miocene and younger oblique thrust or reverse faults on the major structural and topographic features visible on the both sides of the basin (San Juan-Azua, Enriquillo cross sections are the result of both early Miocene oblique valleys). collision between the oceanic plateau of the Presqu'ile du Sud-Bahuroco terrane and the central and northern part of (3) Island arc terranes of the northern and central part of the the island, and the continuing transpressional strike-slip island are topographically high-standing and deeply faulting across Hispaniola. Structural expressions of postu- eroded; the oceanic plateau terrane of the southern part

140 G. DRAPER, P. MANN and J.F. LEWIS

of the island is relatively low-standing and less deeply graphic relief and less extensive evidence of recent faulting. 7 eroded. Bourdon mapped large folds and faults in Cretaceous to Middle Eocene strata which we interpret as a thin-skinned, (4) The prominent (Eocene) folding and thrusting event in northeast-verging fold and thrust belt (Fig. 7.5 D-D'). Bour- eastern Hispaniola verges northeast7. don7 constrained the age of deformations as middle Eocene or younger, because the youngest deformed rocks are of Interpretation of the regional scale structure of middle Eocene age. This Eocene deformation may correlate Hispaniola with late Eocene deformation in the Cordillera Septentrional We propose that island arc and back-arc basin terranes described by de Zoeten and Mann24, which is interpreted as of northern and central Hispaniola form the hanging wall of the result of attempted subduction of the southern edge of a large thrust plate or 'crustal flake' which obliquely over- the Bahama platform beneath eastern and northern His- thrusts the oceanic plateau terrane of southwestern His- paniola (see below). paniola (Fig. 7.5). Matsumoto et al.69 and McCann and 76 Sykes described intermediate depth earthquakes below TECTONIC PHASES AFFECTING TERRANES OF the topographically highest part of the Cordillera Central HISPANIOLA which we suggest may be a manifestation of northeastward underthrusting of the Selle-Hotte-Bahoruco terrane beneath Stratigraphic and structural correlation between the eleven the Cordillera Central. island arc terranes of northern Hispaniola suggest four to This geometry is consistent with the high elevation and five main tectonic phases. topographic asymmetry of the Massif du Nord-Cordillera Central range, and the topographic and structural asymme- Early Cretaceous-middle Eocene arc construction phase try of the Sierra de Neiba. We consider that the major The arc construction phase is recorded by the stratigra- structural manifestation of this tectonic activity is the fault phy of terranes exposed over the northern two-thirds of the which outcrops at the southern edge of the Sierra de Neiba island of Hispaniola. The basal part of the island arc se- range. This fault both juxtaposes island arc and oceanic quence is represented by the Duarte terrane32,33,59. Draper plateau terranes, and exhibits the greatest vertical displace- and Lewis32,33 interpreted the Duarte terrane as a seamount ment of any fault on the island. Furthermore, we suggest that structure and suggested that it is the preserved portion of the the southeast verging thrusts of the southern boundaries of seamount-ocean floor basement upon which the island arc the Presqu'ile du Nord-Neiba and Trois Rivieres-Peralta edifice of Hispaniola was built. That the arc was built on an terranes root in subhorizontal decollements that are at depths elevated (that is, shallower than abyssal ocean floor) sub- of 10 to 15 km. strate is supported by the emergent nature of the upper part Northeast-verging thrust and reverse faults along the of the 3000 m thick Los Ranchos Formation. The Upper northern edge of the Sierra de Bahoruco (Bahoruco fault Cretaceous to Middle Eocene calc-alkaline volcanic and zone) and the northern edge of the Sierra de Neiba (North related rocks of the Amina-Maimon-Tortue, Seibo, Oro and Sierra de Neiba fault zone) are interpreted as back-thrusts. Tireo terranes are thus interpreted as volcanic and sedimen- A minimum age of thrusting is given approximately by the tary accumulations above the older, oceanic Duarte terrane. middle to late Miocene age of the oldest clastic rocks in the There is some evidence for a regional tectonic event or ramp basins of central Hispaniola. events in the early or middle Cretaceous which affected the The southern edge of the Cibao Basin, adjacent to the island arc terranes of central and eastern Hispaniola (Fig. northern edge of the Cordillera Central, appears to be a 7.4). However, the overprinting effects of Cenozoic colli- flexural boundary and this observation is consistent with the sional and strike-slip deformation obscure many of the proposed northeastward underthrusting of the Cordillera manifestations of earlier deformatonal events. Two possi- Central beneath the terranes of northern Hispaniola. In other bilities have been suggested, both of which invoke apolarity words, uplift of the southern edge of the Cordillera Septen- reversal of the arc. trional and subsidence of the Cibao basin appears to be ?Pre-Aptian event driven by transpressional fault movement along the Septen- This event is documented in the Seibo terrane by a trional fault zone24. The early Miocene angular unconfor- major erosional unconformity that separates the Los Ran- mity along the southern flank of the basin dates the initial chos Formation from the overlying shallow-water, Aptian- compressional loading of the basin. Albian Hatillo limestone9. Calc -alkaline andesites, tuffs, The structure and topography of the eastern part of the and limestones (Las Lagunas Formation of Bowin9; Loma island contrasts markedly with that of western and central La Veca volcanics of Lebron55) overlie the limestone. Le- Hispaniola. Eastern Hispaniola exhibits much less topo- bron interpreted the unconformity and accompanying

141 Hispaniola

Figure 7.8. Schematic model for the amalgamation of oceanic plateau and arc terranes to the island arc nucleus of Hispaniola within the broad, 250 km-wide strike-slip plate boundary zone. See text for discussion. transition from primitive island arc to calc -alkaline volcan- Middle Eocene arc collision phase ism as due to a reversal in subduction polarity from north- The collision of the Hispaniola segment of the Greater dipping to south-dipping. Antilles Arc with the Bahama Platform, and subsequent ?Campanian event cessation of north facing subduction, is thought to have Draper and Lewis 33 outlined an alternative model for occurred in the middle to early late Eocene. This age is collision and consequent subduction reversal from a south- inferred from cessation of calc alkaline volc anism in central Hispaniola; uplift and cooling ages of metamorphic rocks of facing to north-facing arc in the Campanian (Fig. 7.6). The 51 effect of this collision would have been to thrust rocks of the the Samana terrane ; age of thrusting in the Seibo and QIC terranes7; and thrusting in the Trois Riviferes-Peralta terrane Tireo and Duarte terranes northeast over rocks of the Loma 25,47,104 Caribe-Tavera and Tortue-Amina-Maimon terranes. North- along the southern flank of the Cordillera Central . ward thrusting produced northeast-verging isoclinal folds In contrast to Cuba, no rocks of the Bahama Platform and regional metamorphism of the Tortue-Amina-Maimon appear to occur in Hispaniola. It is therefore suggested that tenane. Draper and Lewis argued that as the Cenomanian- the southern extension of the platform which collided with Coniacian (97-87 Ma) rocks are involved in the deforma- Hispaniola has a much thinner crust (about 15 km) than the tion, and as late Campanian (80-75 Ma) tonalites are segment that collided with Cuba (20-30 km) and was simply unaffected, that the event must have taken place in the early over-ridden by the Hispaniola arc. Alternatively, the colli- to middle Campanian. This estimate also coincides with the sion may have been highly oblique and did not involve a significant amount of thin-skinned style of folding and ages of blueschists in the northern terranes of Hispaniola 102 The event may correlate with a Campanian deformation and thrusting . uplift event which affected the island arc rocks of central and western Cuba89. This date also coincides with the time Late Eocene-early Miocene strike-slip phase of the initial entry of B" oceanic plateau into the Caribbean Calc-alkaline magmatism terminated after the middle as predicted in some tectonic models87. Eocene collision. Subsequent volcanism on Hispaniola pro-

142 G. DRAPER, P. MANN and J.F. LEWIS

Table 7.1. Correlation of morphotectomc zones and tectonic terranes in Hispaniola. ______

Morphotectonic Zone Tectonic Terrane

Zone 1 Old Bahama Trench (offshore) Zone 2 Cordillera Septentrional-Samana Peninsula Puerto Plata-Pedro Garcia-Rio San Juan; Samana

Zone 3 Cibao Valley One or more of the three following terranes may be in the subsurface of Zone 3: Altamira;

Tortue-Amina-Maimon; Seibo Zone 4 Massif du Nord-Cordillera Central Tortue-Amina-Maimon; Loma Caribe- Tavera; Duarte; Tireo; Trois Rivieres- Peralta Zone 5 Northwestern-south-central zone (includes Presqu'ile du Nord-Ouest-Neiba Plateau Central, San Juan Valley, Azua Plain, Sierra de Ocoa, Presqu'ile du Nord-Ouest) Zone 6 Cul-de-Sac Plain; Enriquillo Valley Selle-Hotte-Bahoruco terrane appears to underlie most of the subsurface of Zone6 Zone 7 Southern Peninsula; Massif de la Selle; Selle-Hotte-Bahoruco Massif de la Hotte; Sierra de Bahoruco

Zone 8 Eastern Peninsula; Cordillera Oriental; Seibo Seibo; Oro coastal plain Zone 9 San Pedro basin and north slope of the One or more of the three following Muertos Trench terranes may be in the subsurface of the San Pedro basin: Loma Caribe-Tavera; Tortue-Amina-Maimon; Seibo

Zone 10 Beata Ridge and Southern Peninsula Selle-Hotte-Bahoruco

duced scattered occurrences of alkalic and calc -alkalic vol- Early Miocene-Recent transpressional phase canic flows and plugs that appear to be related to movement The collision of north-central and southern Hispaniola 98 along strike-slip or strike-slip-related secondary faults . occurred in the early Miocene47, culminating in major uplift Evidence for strike-slip offset across terrane boundaries in and erosion across Hispaniola by the late middle Mio- Hispaniola is mainly indirect or based on regional con- cene24,37,77. This suturing event initiated the transpressional straints: the Cayman Trough, a major pull-apart basin along phase of the island's history which has culminated in the the extension of strike-slip faults passing through His- present-day morphotectonic units of the island (Fig. 7.3A). 90 paniola, exhibits at least 1100 km of left-lateral offset ; Pleistocene indentation of the southern margin of Hispaniola terrane boundaries in Hispaniola are linear fault zones which by the Beata Ridge has contributed to the recent deformation juxtapose rocks of widely varying composition and meta- and uplift of central Hispaniola47. morphic grade (Fig. 7.3B33); and elongate late Eocene-Oli- gocene clastic sedimentary basins in Hispaniola appear to have formed along strike-slip faults25. Several tectonic TECTONIC EVENTS AFFECTING THE models have been proposed which attempt to explain how OCEANIC PLATEAU TERRANE OF large-offset strike-slip offset faults affected the Cretaceous - SOUTHERN HISPANIOLA Eocene island arc and oceanic plateau rocks of His- 14,29,87,95 The tectonic history of the Selle-Hotte-Bahoruco oceanic paniola . plateau terrane of southern Hispaniola was quite distinct from the tectonic history of the island arc terranes of north-

143 Hispaniola ern and central Hispaniolauntil the early Miocene, when the model involves the accretion or welding of forearc/accre- two terranes were sutured together. We recognize four major tionary prism terranes of northern Hispaniola and an oceanic tectonic events in the history of the Selle-Hotte-Bahoruco plateau terrane of southern Hispaniola to a magmatic island terrane. arc nucleus in central Hispaniola The timing of accretionary events can be established by three main criteria: overlap Late Cretaceous oceanic plateau construction phase assemblages that depositionally overlie two distinctive, jux- Radiometric and fossil dates of the Selle-Hotte-Baho- taposed stratigraphic sequences; the sudden appearance of ruco oceanic plateau terrane of southern Hispaniola suggest a detritus in one terrane derived from a dissimilar neighbour; Santonian-Campanian age of formation. The geochemistry and welding together of unlike sequences by intrusions. of the igneous rocks suggest they are similar to basalts From the terrane correlation chart (Fig. 7.7), it can be drilled on the Caribbean seafloor and formed in a hotspot or seen from overlap assemblages that terrane amalgamation oceanic environment94. The age of the oceanic crust beneath the mainly occurred in the Cenozoic with most amalgamation Upper Cretaceous basalt flows in Hispaniola and on the occurring in Miocene to Recent time. Key stratigraphic Caribbean seafloor is unknown, but may be much older relationships which document the age of terrane accretion Jurassic crust12,41. include: (1) late Cretaceous and Paleocene-middle Eocene clastic overlap assemblages of the Trois Rivieres-Peralta Latest Cretaceous deformation and uplift terrane that depositionally overlie the Tireo terrane2,58; (2) A major erosional unconformity above Maastrichtian possible welding of the Duarte and Tireo terranes by the late and older Cretaceous rocks marks the end of significant Cretaceous Loma de Cabrera batholith (Fig. 7.221); (3) igneous activity on the Selle-Hotte-Bahoruco terrane. De- welding of the Tortue-Maimon-Amina and Seibo terranes no formed rocks beneath the unconformity record northward later that Eocene, as demonstrated by an Eocene pluton overthrusting and slumping99. This deformation may be intruded into the terrane boundary9; (4) a late Oligocene related to collision of the terrane with Central America prior to carbonate overlap assemblage (Velazquitos Formation of large-scale left-lateral offset across the northern Carib- the Loma Caribe-Tavera terrane) that depositionally overlies bean87. the Duarte terrane25,84; (5) a late Miocene clastic overlap sequence (Trinchera Formation) derived from the Tireo Paleocene-early Miocene subsidence and strike-slip tenane that depositionally overlies the Presqu'ile du Nord- faulting Ouest-Neiba terrane and the Selle-Hotte-Bahoruco ter - 77,92 The Paleocene-early Miocene history of the Hotte- rane ; (6) a late Miocene clastic overlap sequence Selle-Bahoruco terrane is characterized by strike-slip fault- (Bulla/Cercado Formation) derived from the Duarte and ing, accompanying carbonate sedimentation and scattered Loma Caribe-Tavera terranes that depositionally overlies 84 occurrences of alkaline volcanism (Fig. 7.43). the Amina terrane and some fault strands of the Hispaniola 23,25 fault zone ; (7) a late Miocene-early Pliocene shallow - Early Miocene-Recent transpressional phase water carbonate overlap sequence (Villa Trina Formation) Carbonate sedimentation terminated in late Miocene to that depositionally overlies three terranes along the north Pliocene and post-dated the end of the early to middle coast of the island (Samana, Rio San Juan/Puerto Plata/Pedro Miocene carbonate sedimentation on the Presqu'ile de Garcia, Altamira); and (8) welding of the Selle-Hotte- Nord-Ouest-Neiba terrane to the northwest (Fig. 7.4). Re- Bahoruco, Trois Rivieres-Peralta and Tireo terranes by gional folding of the southern terrane accompanied uplift Plio-Pleistocene basaltic intrusions related to strike-slip 67,98 and erosion and was coeval with folding of the terrane to the movement . north. By the late Miocene, clastic sediments derived from These relationships suggest that the older amalgama- the erosion of the uplifted Tireo, Trois Riv&res-Peralta and tion events occurred closer to the magmatic island arc nu- Presqu'ile du Nord-Ouest island arc terranes had onlapped cleus of central Hispaniola (for example, Cretaceous the Selle-Hotte-Bahoruco terrane, and signified the docking of welding of Duarte and Tireo terranes; Oligocene overlap the island arc terranes and the oceanic plateau77. between the Loma Caribe-Tavera and Duarte terranes) aid the younger accretion events involving the oceanic plateau DISCUSSION of southern Hispaniola and forearc/accretionary prism terranes of northern Hispaniola occurred in a more outboard Model for terrane amalgamation in Hispaniola position (Fig. 7.8). Because lateral movements associated Based on the terrane correlation chart of Figure 7.7, we with North America-Caribbean motion were at a low angle present a model for terrane amalgamation, or the complete to the 'structural grain’ or strike, the extinct island arc assembly history of terranes in Hispaniola (Fig. 7.8). This structure approximately maintained its original cross sec -

144 G. DRAPER, P. MANN and J.F. LEWIS

tional profile with forearc/accretionary prism terranes in Eocene. The terranes can be classified variously as: (1) northern Hispaniola, a magmatic nucleus in central His- fragments of the forearc/accretionary prism of an island paniola, and a back-arc basin in southern and central His- arc; (2) fragments of the magmatic part of an island arc; paniola. More detailed studies of offsets across major (3) fragment of a back-arc basin; and (4) fragments of Neogene faults are needed before realistic pre-Eocene cross oceanic crust. sections of the Hispaniola island arc can be drawn. (2) The structure and stratigraphy of the island arc terranes record four main tectonic phases: (1) early Cretaceous Terrane dispersal versus terrane accretion and the to middle Eocene island arc volcanism magmatism (this northern Caribbean offset problem arc construction phase may have been interrupted by Strike-slip faulting can produce terrane dispersion, or one or two poorly known deformation events in the the breakup of larger terranes into smaller piec es49. The late pre-Aptian and Campanian); (2) middle-late Eocene Eocene and younger tectonic history of Hispaniola appears to collision and uplift of the arc with the southern edge of be an example of strike-slip faulting associated with the North America plate (southeastern Bahama carbon simultaneous terrane accretion and dispersion29. Our expla- ate platform) resulting in the abrupt termination of nation for this phenomenon lies in the fact that the island arc island arc activity; (3) late Eocene to early Miocene nuc leus of Hispaniola has constituted a restraining bend east-west strike-slip faulting at a low angle to the strike segment in a major transcurrent plate boundary zone65. The of the extinct island arc; and (4) early Miocene tran- initiation of the restraining bend seems to coincide with the spression resulting from oblique collision ('docking') early Miocene suturing of the oceanic plateau of southern of the southern oceanic plateau terrane with the nucleus Hispaniola to the island arc nucleus of central Hispaniola of Hispaniola. From this time, the bend area was progressively widened by (3) The oceanic plateau terrane of southwestern Hispaniola terrane accretion, probably at the expense of terrane 'disper- experienced four main tectonic phases, the first three of sion' of along-strike segments of the island arc in the Cuba which are distinct from the rest of Hispaniola: (1) and Puerto Rico-Virgin Islands portions of the island arc formation in an intraplate hotspot or oceanic island chain (Fig. 7.8). setting from at least the Santonian until the Maas- The model of terrane amalgamation at a strike-slip trichtian.; (2) Maastrichtian deformation, rapid uplift restraining bend helps to explain the ‘northern Caribbean and erosion; (3) strike-slip faulting, subsidence, and offset problem', or the discrepancy between the 1100 km formation of a carbonate platform and basin from Pa- offset caused by the opening of the Cayman Trough and the leocene through early Miocene time; and (4) transpres- apparently much smaller offsets seen in on-land faults in sion resulting from early Miocene oblique collision Central America and the Greater Antilles13,15,30,63. We with the island arc terranes of central Hispaniola. suggest that the offset discrepancy may reflect the transfor- (4) Some of the terrane boundaries separating island arc and mation of a large strike-slip offset along straight faults in the oceanic plateau terranes were reactivated or obliquely Cayman Trough to smaller strike-slip offsets with greater cut by large oblique-slip faults during the early Miocene vertical motions along curved, restraining bend fault seg- to Recent convergence between the island arc and the ments in the Central America and Hispaniola The offset in southern oceanic plateau terranes to form nine mor- the restraining bend areas is absorbed by either by under - photectonic provinces consisting of elongate, fault- thrusting of one terrane beneath another (Selle-Hotte-Baho- bounded mountain ranges and ramp basins. ruco terrane beneath island arc terranes of central ACKNOWLEDGEMENTS—The authors acknowledge continued fiscal Hispaniola), or by large-scale rotation of terranes about support from their respective institutions in their field investigations in vertical axes such as the post-Eocene counterclockwise Hispaniola. Substantial grants were also received from other organizations rotation of the Tireo terrane97, and/or by splaying of the (National Science Fountation EAR-83061452, EAR-8509542, Latin strik e-slip faults into several different strands at the restraining American and Caribbean Center of FIU grants to Draper; National Science Foundation EAR-8608832, Petroleum Research Fund of the American bends with small amounts of offset being accommodated by Chemical Society 17068-AC2, Tenneco and Pecten International to Mann; 87 each fault splay . Additional offset may be absorbed on National Science Foundation and the United Nations Revolving Fund for unrecognized strike- or oblique-slip faults, or by internal Natural Resources Exploration to Lewis). We greatly appreciate the con- deformation of fault blocks. tinued cooperation and field assistance of government and university officials in the Dominican Republic and Haiti. SUMMARY AND CONCLUSIONS

(1) Hispaniola consists of an amalgamation of twelve terranes which range in age from Early Cretaceous to late

145 Hispaniola

REFERENCES Conference, Santo Domingo, Dominican Republic, 16th-20th August, 1980, 39-50. 1Bellon, H., Vila, J.-M. & Mercier de Lepinay, B. 1985. 12Bowland C.L. & Rosencrantz, E. 1988. Upper crustal Chronologic 40K-39Ar et affinites des manifestations structure of the western Colombian Basin, Caribbean magmatiques ail Cretace et au Paleogene dans Tile Sea. Geological Society of America Bulletin, 100, 534- d'Hispaniola: in Mascle, A. & Biju-Duval, B. (eds), 546. Geodynamique des Caribes. Editions Technip, Paris, 13Burkart, B., Deaton, B.C., Dengo, C. & Moreno, G. 1987. 329-340. Tectonic wedges and offset Laramide structures along 2Biju-Duval, B., Bizon, G., Mascle, A. & Muller, C. 1983. the Polochic fault of Guatemala and Chiapas, Mexico: Active margin processes: field observations in southern reaffirmation of large Neogene displacement. Tecton- Hispaniola. American Association of Petroleum Geolo- ics, 6, 411-422. gists Memoir, 34, 325-346. 14Burke, K. 1988. Tectonic evolution of the Caribbean. 3Bizon, G. Bizon, J.J., Calmus, T., Miiller, C. & van den Annual Review of Earth and Planetary Sciences, 16, Berghe, B. 1985. Stratigraphie du Tertiaire du sud 201-230. d'Hispaniola (Grandes Antilles): influence de la tec- 15Calais, E. & de Lepinay, B.M. 1990. Tectonique et tonique decrochante sur lapaleogeographie et 1'histoire paleogeographie de la cote sud de I'Oriente cubain: sedimentaire: in Mascle, A. & Biju-Duval, B. (eds), nouvelles contraintes pour 1'evolution geodynamique Geodynamique des Caribes. Editions Technip, Paris, de la limite de plaques decrochante nore-caraibe de 371-380. 1'Eocene a 1'actuel. Comptes Rendus de la Academie 4Bohor, B. & Sietz, R. 1990. Cuban K/T catastrophe. Na- des Sciences, 310, 293-299. ture, 344, 593. 16Calmus, T. 1983. Contribution a Vetude geologique du 5Boisseau, M. 1987. Leflac nord-est de la Cordillere Central massif de Macaya (sud-ouest d'Haiti, Grandes An- dominicaine (Hispaniola, Grandes Antilles): un edifice tilles): sa place dans I 'evolution de I 'orogene nord- de Nappes Cretace polyphase. Unpublished Ph.D. caraibe. Unpublished Ph.D. thesis, Universite Pierre et thesis, Universite Pierre y Marie Curie, Paris. Marie Curie, Paris. 6Boisson, D. 1987. Etude geologique du Massif du Nord 17Calmus, T. 1987. Etudes geochimiques des volcanismes d'Haiti (Hispaniola-Grandes Antilles). Unpublished Cretace et Tertiaire du massif de Macaya (presqu'ile du Ph.D. thesis (3rd cycle), Universite Marie et Pierre Sud d'Haiti): leur place dans 1'evolution tectonique de Curie, Paris. laregionnord-caraibe. Comptes Rendus de la Academie 7Bourdon, L. 1985. La Cordillere Orientale Dominicaine des Sciences, 304, 981-986. (Hispaniola, Grandes Antilles): un arc insulaire Cretace 18Case, I.E., Holcombe, T.L. & Martin, R.G. 1984. Map of polystructure. Unpublished Ph.D. thesis, Universite geologic provinces in the Caribbean region. Geological Pierre et Marie Curie, Paris. Society of America Memoir, 162,1-30. 8Bourdon, L., Geyssant, J.R., Mercier de Lepinay, B. & 19Cheilletz, A., Kachrillo, J. J., Sonet, J. & Zimmerman, J.L., Vila, J.-M. 1984. Interets paleontologique, chronolo- 1978. Petrographie et geochronologie de deux com- gique et tectonique de la decouverte de Peroniceras plexes intrusifs aporphyres cupriferes d'Haiti. Contri- (Ammonoidea, Collignoniceratidae) dans le Coniacien bution a la connaiscance de la province cuprifere inferieur de la Cordillere orientale de Republique laramienne de 1'arc insulaire de Grande Antilles. Bul- Dominicaine (Hispaniola, Grandes Antilles). Comptes letin de la Societe Geologique de France, series 7, 20, Rendus de Academie des Sciences, 298, 287-292. 107-114. 9Bowin, C.O. 1966. Geology of the central Dominican 20Cooper, J.C. 1983. Geology of the Fondo Negro region, Republic —& case history of part of an island arc. Geo- Dominican Republic. Unpublished M.S. thesis, State logical Society of America Memoir, 98, 11-84. University of New York, Albany. 10Bowin, C.O. 1975. The geology of Hispaniola: in Nairn, 21Cribb, J.W., Jimenez, J., Lewis, J.F. & Sutter, J. F. 1989. A.E.M. & Stehli, F.G. (eds), The Ocean Basins and 40Ar/39Ar ages from Loma de Cabrera batholith: impli- Margins. 3. The Gulf of Mexico and the Caribbean. cations for timing of tectonic events in northern His- Plenum, New York, 500-552. paniola. Geological Society of America Abstracts with 11Bowin, C.O. & Nagle, F. 1982. Igneous and metamorphic Programs, 21 (7), A267. rocks of northern Dominican Republic: an uplifted 22de Lepinay, B.M., Labesse, B., Sigal, J. & Vila, J.M. 1979. subduction zone complex: in Snow, W., Gil, N., Llinas, Sedimentation chaotique et tectonique tangentielle R., Rodriguez-Torres, R., Seaward, M. & Tavares, I. Maestrichtiennes dans la presqu'ile du sud d'Haiti (lie (eds), Transactions of the 9th Caribbean Geological d'Hispaniola, Grandes Antilles). Comptes Rendus de la

146 G. DRAPER, P. MANN and J.F. LEWIS

Academie des Sciences Paris, series D, 187, 887-890. cial Paper, 262, 29-45. 23de Riemer, V.W. 1978. Results of geological investiga- 34Draper, G. & Nagle, F. 1988. Geological setting and tions in the northwestern part of the Dominican Repub- characteristics of blueschist and eclogite bearing me- lic. Neues Jahrbuch Jur Geologie und Paleontologie langes in northern Hispaniola: in Barker, L. (ed.), Monatshefte, 35,162-172. Transactions of the Eleventh Caribbean Geological 24de Zoeten, R. & Mann, P. 1991. Structural geology and Conference, Dover Beach, Barbados, 20th-26th July, Cenozoic tectonic history of the central Cordillera Sep- 1986,33:1-33:9. tentrional, Dominican Republic. Geological Society of 35Draper, G. & Nagle, F. 1991. Geology, structure and America Special Paper, 262,265-279. tectonic development of the Rio San Juan Complex, 25Dolan, J., Mann, P., de Zoeten, R., Heubeck, C., Shiroma northern Dominican Republic. Geological Society of & Monechi, S. 1991. Sedimentologic, stratigraphic, America Special Paper, 262, 77-95. and tectonic synthesis of Eocene-Miocene sedimentary 36Eberle, W., Hirdes, W., Muff, R. & Peaez, M. 1982. The basins, Hispaniola and Puerto Rico. Geological Society geology of the Cordillera Septentrional (Dominican of America Special Paper, 262, 217-263. Republic): in Snow, W., Gil, N., Llinas, R., Rodriguez- 26Donnelly, T.W. 1989. Geologic history of the Caribbean Torres, R., Seaward, M. & Tavares, I. (eds), Transac- and Central America: in Bally, A.W. & Palmer, A.R. tions of the Ninth Caribbean Geological Conference, (eds), The Geology of North America. Volume A. An Santo Domingo, Dominican Republic, 16th-20th Au- Overview, 299-321. Geological Society of America, gust, 1980, 619-632. Boulder. 37Edgar, N.T. 1991. Structure and geologic development of the 27Donnelly, T. W., Beets, D., Carr, M. J., Jackson, T., Klaver, Cibao Valley, northern Hispaniola. Geological Society of G., Lewis, J., Maury, R., Shellenkens, H., Smith, A.L., America Special Paper, 262, 281-299. Wadge, G. & Westercamp, D. 1990. History and tec- 38Edgar, N.T. et al 1973. Initial Reports of the Deep Sea tonic setting of Caribbean magmatism: in Dengo, G. & Drilling Project, Volume XV. U.S. Government Printing Case, J.E. (eds), The Geology of North America. Vol- Office, Washington, D.C., 1137 pp. ume H. The Caribbean Region, 339-374. Geological 39Espaillat, J., Bloise, G., Lewis, J.F. & Macveigh, J.G. Society of America, Boulder. 1990. Petrography and geochemistry of mafic rocks of 28Donnelly, T.W. & Rogers, J.J.W. 1978. The distribution the Peralvillo Formation in the Sabana Potrero area, of igneous rock suites throughout the Caribbean. central Dominican Republic: in Larue, D.K. & Draper, G. Geologie en Mijnbouw, 57,151-162. (eds), Transactions of the Twelth Caribbean Geological 29 Draper, G. 1989. Terrane accretion and re-shuffling in Conference, St. Croix, Virgin Islands, 7th-llth August, Hispaniola. Geological Society of America Abstracts 1989,190-199. with Programs, South-Central Section, 21, 9. 40Garcia, E. & Haims, F. 1989. Informe del mapageologico de 30Draper, G. & Barros, J.A. 1988. Tectonic reconstruction la Republica Dominicana, San Juan (5972). Secre-taria of N. Hispaniola and S.E. Cuba: dissection of an de Industria y Comercio, Direction General de Mineria, Cretaceous island arc. Geological Society of America Santo Domingo, 97 pp. Abstracts with Programs, 20 (7), A60. 41Ghosh, N., Hall, S.A. & Casey J.F. 1984. Seafloor spreading 31 Draper, G. & Lewis, J.F. 1982. Petrology, deformation, magnetic anomalies in the Venezeulan Basin. Geological and tectonic significance of the Amina Schist, northern Society of America Memoir, 162, 65-80. Dominican Republic: in Snow, W., Gil, N., Llinas, R., 42Girard, D., Beck, C., Stephan, J.F., Blanchet, R. & Maury, R. Rodriguez-Torres, R., Seaward, M. & Tavares, I. 1982. Petrologie, geochimie et signification geodyn- (eds), Transactions of the Ninth Caribbean maique de quelques formations volcaniques cretacees Geological Conference, Santo Domingo, Dominican peri-caraibes. Bulletin de la Societe Geologique de Republic, 16th-20th August, 1980, 53-64. France, series 7, 24,535-544. 32 Draper, G. & Lewis, J.F. 1989. Petrology and structural 43Goreau, P. 1981. The tectonic evolution of the north development of the Duarte complex, central Dominican central Caribbean plate margin. Unpublished Ph.D. Republic: a preliminary account and some tectonic thesis, Massachusetts Institute of Technology. implications: in Duque-Caro, H. (ed.), Transactions of 44Groetsch, G.J. 1982. Resedimented conglomerates and the Tenth Caribbean Geological Conference, turbidites of the Upper Tavera Group, Dominican Re- Cartagena, Colombia, 14th-20th August, 1983, 103- public: in Snow, W., Gil,N.,Llinas, R.,Rodriguez-Torres, 112. R., Seaward, M. & Tavares, I. (eds), Transactions of the 33 Draper, G. & Lewis, J.F. 1991. Metamorphic belts in Ninth Caribbean Geological Conference, Santo central Hispaniola. Geological Society of America Spe- Domingo, Dominican Republic, 16th-20th August,

147 Hispaniola

1980, 191-198. Volume H. The Caribbean Region, 77-140. Geological 45Haldemann, E.G., Brouwer, S.B., Blowes, J.H. & Snow, Society of America, Boulder. 58 W. E. 1980. Lateritic nickel deposits at Bonao Falcon- Lewis, J.F., Amarante, A., Blaise, G., Jimenez, G. & bridge Dominicana C. por A. Field Guide, Ninth Car- Dominguez, H.D. 1991. Lithology and stratigraphy of ibbean Geological Conference, Santo Domingo, Upper Cretaceous volcanic and volcaniclastic rocks of Dominican Republic, 69-78. the Tireo Group, Dominican Republic, and correlations 46Heubeck, C. 1988. Geology of the southeastern termina- with the Massif du Nord in Haiti. Geological Society of tion of the Cordillera Central, Dominican Republic, America Special Paper, 262,143-163. 59 Greater Antilles. Unpublished M.A. thesis, Lewis, J.F. & Jimenez, J.G. 1991. Duarte Complex in the University of Texas at Austin. La Vega-Janico area, central Hispaniola geological and 47Heubeck, C., Mann, P., Dolan, J. & Monechi, S. 1991. geochemical features of the sea floor during early stages Diachronous uplift and recycling of sedimentary basins of arc evolution. Geological Society of America Special during Cenozoic tectonic transpression, Hispaniola Paper, 262, 115-141. 60 Sedimentary Geology, 7,1-32. Lewis, J.F., Vespucci, P., Robinson, E., Jiang, M. & 48Hildebrand, A.R. & Boynton, W.V. 1990. Proximal Cre- Bryant, A. 1989. Paleogene stratigraphy of the Padre taceous-Tertiary boundary impact deposits in the Car- Las Casas and adjacent areas in the southeast Cordillera ibbean. Science, 248, 843-847. Central, Dominican Republic: in Duque-Caro, H. (ed.), 49Howell, D., Jones, D. & Schermer, E. 1985. Tectonostra- Transactions of the Tenth Caribbean Geological Con- tigraphic terranes of the circum-Pacific region: in How- ference, Cartagena, Colombia, 14th-20th August, ell, D. (ed.), Tectonostratigraphic terranes of the 1983, 229-237. 61 Circum-Pacific. Circum-Pacific Council for Energy Lidz, B. & Nagle, F. (eds). 1979. Hispaniola: tectonic and Mineral Resources, Earth Science Series, 1, 3-30. focal point of the northern Caribbean. Miami Geologi- 50Joyce, J. 1985. High pressure-low temperature metamor- cal Society, Miami, 96p. 62 phism and tectonic evolution of the Samana Peninsula, Llinas, R.A. 1972. Geologia del area Polo-Duverge, Dominican Republic (Greater Antilles). Unpublished Cuenca de Enriquillo. Publication of Colegio Domini- Ph.D. thesis, Northwestern University. cana de Ingenieros, Arquitectosa y Agrimensores, 51Joyce, J. 1991. Blueschist metamorphism and deformation Codia. Part 1 in no. 3,55-65; part 2 in no. 32, 40-53. 63 of the Samana peninsula—a record of subduction and Malfait, B.T. & Dinkelman, M.G. 1972, Circum-Carib- collision in the Greater Antilles. Geological Society of bean tectonic and igneous activity and the evolution of America Special Paper, 262, 47-76. the Caribbean plate. Geological Society of America 52Kesler, S.E. 1971. Petrology of the Terre-Neuve Igneous Bulletin, 83, 250-272. province, northern Haiti. Geological Society of 'America 64Mann, P. & Burke, K. 1984. Neotectonics of the Carib- Memoir, 130, 119-137. bean. Reviews of Geophysics and Space Physics, 22, 53Kesler, S.E., Russell, N., Reyes, C. Santos, L., Rodriguez, 309-362. A. & Fondeur, L. 199la. Geology of the Maimom 65Mann, P., Burke, K. & Matsumoto, T. 1984. Neotectonics of Formation, Dominican Republic. Geological Society of Hispaniola: plate motion, sedimentation, and seis- America Special Paper, 262, 173-185. micity at a restraining bend. Earth and Planetary Sci- 54Kesler, S.E., Russell, N., Polanco, J., McCurdy, K. & ence Letters, 70, 311-324. Cummin, G.L. 199 Ib. Geology and geochemistry of the 66Mann, P., Draper, G. & Lewis J. F. (eds). 1991. Geologic early Cretaceous Los Ranchos Formation, central Do- and tectonic studies the North American Caribbean minican Republic. Geological Society of America Spe- Plate Boundary in Hispaniola. Geological Society of cial Paper, 262, 187-201. America Special Paper, 262, 401 pp. 55Lebron, M.C. 1989. Petrochemistry and tectonic signifi- 67Mann, P., Draper, G. & Lewis, J.F. 1991. Overview of the cance of late Cretaceous, calc-alkaline volcanic geology of Hispaniola. Geological Society of America rocks, Cordillera Oriental, Dominican Republic. Special Paper, 262, 1-28. Unpublished M.S. thesis, University of Florida. 68Mann, P, Hempton, M., Bradley, D. & Burke, K. 1983. 56Lewis, J.F. 1980. Resume of the geology of Hispaniola. Development of pull-apart basins. Journal of Geology, Field Guide, Ninth Caribbean Geological Conference, 91, 529-554. Santo Domingo, Dominican Republic, 5-31. 69Matsumoto, T., Rabb, L., Perez, M.J., Luciano, F. 57 Lewis, J.F. & Draper, G. 1990. Geology and tectonic Sanchez, J. & Pennington, W. 1981. Seismology in the evolution of the northern Caribbean margin: in Dengo, Dominican Republic as inferred from a local network. G. & Case, J.E. (eds), The Geology of North America. EOS, 62, 324.

148 G. DRAPER, P. MANN and J.F. LEWIS

70Mauirrasse, FJ.-M. 1981. New data on the stratigraphy of Rodriguez-Torres, R., Seaward, M. & Tavares, I. (eds), the southern peninsula of Haiti: in Maurrasse, F. J.-M., Transactions of the Ninth Caribbean Geological Con- ed., Transactions du ler Colloque sur la Geologie ference, Santo Domingo, Dominican Republic, 16th- d'Haiti, Port-au-Prince, Haiti, Ministre des Mines et 20th August, 1980, 409-415. des Resources Energftiques, 184-198. 82Nicolini, P. 1977. Les porphyres cupriferes et les com- 71Maurrasse, F.J.-M.R (ed.). 1981. Transactions du ler plexes ultra-basiques du nord-est d 'Haiti: essai gitolo- colloque sur la geologied'Haiti, Port-au-Prince, Haiti. gie previsionnelle. Unpublished Ph.D. thesis (3rd Ministre des Mines et des Resources Energetiques, cycle), University Pierre et Marie Curie, Paris. Port-au-Prince, 286 pp. 83Nicolini, P. 1981. Gitologie Haitienne: in Maurrasse, F.J.- 72Maurrasse, F.J-M.R. 1990. Stratigraphic correlation for M.R. (ed.), Transactions du ler Colloque sur la Geologie the circum-Caribbean Region: in Dengo, G. & Case, d'Haiti, Port-au-Prince, Haiti, 105-111. Ministre des I.E. (eds), The Geology of North America. Volume H. Mines et des Ressources Energetiques, Port-au- The Caribbean Region, plate 4. Geological Society of Prince. America, Boulder. 84Palmer, H.C. 1979. Geology of the Moncion-Jarabacoa 73Maurrasse, FJ.-M., Husler, J., Georges, G., Schmitt, R. & area, Dominican Republic: in Lidz, B. & Nagle, F. Damond, P. 1979. Upraised Caribbean sea-floor below (eds), Hispaniola: tectonic focal point of the northern acoustic reflector B" at the southern peninsula of Haiti. Caribbean, 29-68. Miami Geological Society, Miami. Geologie en Mijnbouw, 58, 71-83. 85Peralta-Villar, J. 1985. Geologie und Erzjuhrung in der 74Maurrasse, FJ-M.R. & Sen, G. 1991. Impacts, tsunamis Umgebung des Intusionsbrekzienkorpers von Los and the Haitian Cretaceous boundary layer. Science, Jobos/Pedro Garcia, Dominikanische Republik. Un- 252,1690-1693. published Diploma thesis, Mineralogisch-Pet- 75McCann, W.R. & Pennington, W.D. 1990. Seismicity, rographischen Institut der Universitat Heidelberg. large earthquakes, and the margin of the Caribbean 86Perfit, M.R & McCullough, M.T. 1982. Trace element, Plate: in Dengo, G. & Case, I.E. (eds.), The Geology of Nd-Sr isotope geochemistry of eclogites and North America. Volume H. The Caribbean Region, blueschists from the Hispaniola-Puerto Rico subduc- 291-306. Geological Society of North America, Boul- tion zone. Terra Cognita, 2, 321. der. 87Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of 76Mc Cann, W.R. & Sykes, L.R. 1984. Subduction of aseis- the Caribbean region: a plate tectonic perspective: in mic ridges beneath the Caribbean plate; implications Dengo, G. & Case, J.E. (eds), The Geology of North for the tectonics and seismic potential of the north- America. Volume H. The Caribbean region, 405-432. eastern Caribbean. Journal of Geophysical Research, Geological Society of America, Boulder. 89, 4493-4519. 88Pindell, J.L. & Draper, G. 1991. Stratigraphy and geological 77McLaughlin, P.P., van den Bold, W.A. & Mann, P. 1991. history of the Puerto Plata area, northern Dominican Geology of the Azua and Enriquillo basins, Dominican Republic. Geological Society of America Special Pa- Republic, 1: Neogene lithofacies, biostratigraphy, bio- per, 262, 97-114. facies and paleogeography. Geological Society of 89Pszczolkowski, A. & Flores, R. 1986. Fases tectonicas del America Special Paper, 262, 337-366. Cretacico y del Paleogene en Cuba occidental y central. 78Muff, R. & Hernandez, M. 1986. The hydrothermal altera- Bulletin of the Polish Academy of Sciences (Earth Sci- tion and pyrite-galena-sphalerite mineralization of a ences), 34, 95-111. prophyrite intrusion at Palma Picada in the Cordillera 90Rosencrantz, E., Ross, M. & Sclater, J.G. 1988. Age and Septentrional, Dominican Republic. Natural Resources spreading history of the Cayman Trough as determined and Development, 26, 83-94. from depth, heat flow, and magnetic anomalies. Journal 79Nagle, F. 1974. Blueschist, eclogite paired metamorphic of Geophysical Research, 93, 2141-2157. belts and the early tectonic history of Hispaniola. Geo- 91Russell, W. & Kesler, S.E. 1991. Geology of the Duarte logical Society of America Bulletin, 85, 1461-1466. complex hosting precious mineral mineralization at 80Nagle, F. 1979. Geology of the Puerto Plataarea, Domini- Pueblo Viej, Dominican Republic. Geological Society can Republic: in Lidz, B. & Nagle, F. (eds), Hispaniola: of America Special Paper, 262, 203-215. tectonic focal point of the northern Caribbean, 1-28. 92 Ruth, M.D. 1989. Cenozoic geology of the western San Miami Geological Society, Miami. Juan Valley, Dominican Republic. Unpublished M.S. 81Nagle, F., Wassail, H., Tarasiewicz, G. & Tarasiewicz, E. thesis, George Washington University. 1982. Metamorphic rocks and stratigraphy of central 93Sayeed, U., Maurrasse, F., Keil, K., Husler, J. & Schmidt, R. Tortue Island, Haiti: in Snow, W., Gil, N., Llinas, R., 1978. Geochemistry and petrology of some mafic

149 Hispaniola

rocks from Dumisseau, Haiti. EOS, 59, 403. Danien dans le sud d'Haiti (Riviere Gosseline, Grandes 94Sen, G., Dickey-Vargas, R, Waggoner, D.G. & Maur- Antilles): consequences sur F extension et la mise en rasse, F. 1988. Geochemistry of basalts from the Du- place de la nappe de Macaya. Bulletin de la Societe misseau Formation, southern Haiti: implications for the Geologique de France, series 8,8, 349-359. origin of the Caribbean sea crust. Earth and Planetary 100Vila, J.-M., Butterlin, J., Calmus, T., Mercier de Lepinay, Science Letters, 87, 423-437. B. & van den Berghe, B. (compilers!. 1983. Geologic 95Sykes, L.R., McCann, W.R. & Kafka, A.L. 1982. Motion map of Haiti: in Girault, M. (ed.), Atlas d 'Haiti, planche of Caribbean plate during last 7 million years and V, carte geologique, scale 1:1,000,000. implications for earlier Cenozoic movements. Journal 101Vila, J.-M., Butterlin, J., Labesse, B. & Mercier de Lepi- of Geophysical Research, 87, 10656-10676. nay, B. 1984. Donnees nouvelles sur les roches meta- 96Van den Berghe, B. 1983. Evolution sedimentaire etstruc- morphiques de Tile de la Tortue (Hispaniola, Grandes turale depuis le Pal eocene du secteur "Massif de la Antilles). Comptes Rendus de la Academie des Sci- Selle" (Haiti)-"Baoruco" (Republique Dominicaine)- ences, Paris, 294,1103-1106. "Nord de la Ride de Beata" dans Vorogene nord 102Wadge, G., Draper, G. & Lewis, J.F. 1984. Ophiolites of Caraibe (Hispaniola-Grandes Antilles). Unpublished the northern Caribbean: a reapraisal of their roles in the Ph.D. thesis (3rd cycle), University de Paris. evolution of the Caribbean Plate: in Gass, I.G., Lippard, 97van Fossen, M.C. & Channell, J.E.T. 1988. Paleomag- S.I. & Shelton, A,W. (eds), Ophiolites and oceanic netism of late Cretaceous limestones and chalks from lithosphere. Geological SocietjrfLondon Special Pub- Haiti: tectonic interpretations. Tectonics, 7, 601-612. lication, 13, 367-380. 98Vespucci, P. 1989. Summary of trace element and isotope 103Weyl, R. 1966. Geologie der Antillen, Band 4, Beitrage data, late Cretaceous volcanics, Hispaniola: in zur Regionalen Geologie der Erde (HJ. Martini, ed.). Duque-Caro, H. (ed.), Transactions of the Tenth Gebruder Borntraeger, Berlin, 410 pp. Caribbean Geological Conference, Cartagena, 104Witschard, M. & Dolan, J.F. 1990. Contrasting structural Colombia, 14th-20th August, 1983, 467-472. styles in sili clastic and carbonate rocks of an offscraped 99Vila, J-M., Amilcar, H., Amilcar, H.C., Boisson, D. & sequence: the Peralta accretionary prism. Geological Feinberg, H/1990. Un evenement tectono-sedimentaire Society of America Bulletin, 102, 792-806.

CHAPTER 8

Puerto Rico and the Virgin Islands

DAVID K. LARUE

Exxon Production Research Company, P.O. Box 2189, Houston, Texas 77252-2189, U.SA.

INTRODUCTION from the St. Croix bank as young as Oligocene have been reported . THE GEOLOGIC evolution of the following geographic From the Oligocene to the Holocene in Puerto Rico, and regions is discussed herein: the island of Puerto Rico; the from the Miocene to the Holocene in the northern Virgin Puerto Rican Virgin Islands; the northern Virgin Islands; Islands and St. Croix, the region was covered by limestones and St Croix (Fig. 8.1). Oceanic realms considered include and sediments derived from the weathering of the island arc the Puerto Rico and Muertos Trenches, the Virgin Islands massif30,71,85. Complicated tectonism from het middle Mio- Basin, the Mona Passage and the Mona Canyon. cene to the Holocene in Puerto Rico and the Virgin Islands Puerto Rico, the northern Virgin Islands and St. Croix is associated with easterly transtensional motion of the represent the eastern edge of the ancestral Greater Antilles Caribbean Plate relative to North America32, and 25° anti- island arc (Fig. 8.1). Arc volcanism in the eastern Greater clockwise rotation of the Puerto Rico and northern Virgin Antilles, associated with subduction from the north, com- Islands Block67. Extensional structures formed during this menced in the Cretaceous and continued until the Eocene, time include the western Puerto Rico Trench, the Virgin extending into the Oligocene in the northern Virgin Is- 1 Island Basin, Anegada Passage, the 19° Latitude Fault, the lands ' : arc volcanism may have ceased in the late Cre- Mona Canyon and the Mona Allochthon. Thrusting also taceous or Oligocene on St. Croix, based on conflicting occurred in the Muertos Trench and the eastern Puerto Rico interpretations of the K- AT ages. Cessation of arc volcanism Trench during this time period. in the Tertiary was accompanied by orogenesis, including Owing to the complicated tectonic evolution of the area, folding, faulting and local uplift on the order of several successive deformations and evolutionary stages are de- kilometres. It is believed that orogenesis and termination of scribed herein from youngest to oldest, in three parts, in arc volcanism in the early Tertiary was related to either the order that the reader will appreciate better the geologic collision of the western (Cuban) part of the Greater Antilles evolution of the region. arc with the Bahama Bank , subduction of the Bahama Bank beneath the Puerto Rican part of the arc22, or was 46 associated with subduction of buoyant oceanic crust . On PART 1: NEOTECTONICS Puerto Rico, this early Tertiary orogenic event is dated by a late Eocene to middle Oligocene angular unconformity; in Seven principal tectonic units are recognised in the Puerto the northern Virgin Islands, this unconformity ranges in age Rico-Virgin Islands region: Atlantic Oceanic crust; Vene- 1330 from late Oligocene to late Miocene ' . A transitional zuelan Basin crust; the Lesser Antilles island arc; the Puerto period prior to main orogeny is recognised on Puerto Rico Rico-northern Virgin Islands (PRNVI) Block; the St. Croix from the Paleocene to the Eocene, as indicated by zones of Block; the Muertos accretionary complex; and the north 74 localized uplift and subsidence. On St. Croix, Speed et al. slope complex (Fig. 8.2). The northern part of the Lesser argued that orogenesis occurred in the late Cretaceous be- Antilles arc is a complex feature, broken itself into a number cause undeformed plutons of this antiquity cut deformed of structures such as the Saba, Anguilla and Antigua-Bar- rocks. More recent research has recognised the presence of buda Blocks. The geology of the Atlantic and Venezuelan 19 deformed plutons and whole-rock volcanic ages of rocks Basin crusts are discussed by Donnelly (Chapter 3, herein).

151 DAVID K.LARUE

Figure 8.1. Location map of Puerto Rico and the Virgin Islands. Numbers 1-7 are drillholes: 1,2 and 3 are CPR 1- 3 drilled onshore of southern Puerto Rico in 1959-1960; 4 is CPR-4, drilled on the north coast of Puerto Rico in 1960; 5 is the Ram Head borehole drilled on St. John17; 6 is the Turtlehead well, drilled offshore of Tortola by Mobil Exploration and Production, Inc.; 7 is the Toa Baja drill site, drilled on the north coast of Puerto Rico in 1989 by the Puerto Rican Electric Power Authority44. Key: A.=Anegada; C.=Culebra; M.C.=Mona Canyon; S.C.=St. Croix; S.J.=St. John; S.T =St. Thomas ; T =Tortola; V =Vieques; V.G.=Virgin Gorda; V.I. Basin= Virgin Island Basin.

For the purposes of the current discussion, it is relevant to studies associated with this slab in the vicinity of Puerto note that the Atlantic crust consists of relatively normal Rico indicate motions are nearly parallel to the strike of the oceanic crust, whereas the Venezuelan Basin crust is thick inclined seismic zone58. Studies of deformation in the oceanic crust63. Puerto Rico Trench using seismic reflection and GLORIA The PRNVI Block and the St. Croix Block are separated side-scan data indicate thrusting is more prevalent in the by the Virgin Islands Basin and the Anegada Passage31. east20,46,53. Both blocks are considered to be internally rigid, or not The South Samana Bay and Septentrional Faults of possessing any active large magnitude or offset faults, and Hispaniola52 can be traced using bathymetry toward the east are under lain by approximately 25-30 km of island arc where they coincide with the Puerto Rico Escarpment (Fig. crust63,77. The PRNVI Block is clearly a composite terrene 8.1). Larue et al.46 have argued that two nearly-coincident comprised of several accreted fragments of island arc. Both faults exist near the Puerto Rico Escarpment, defining a blocks have subducted Atlantic oceanic crust beneath them narrow horst: a south-dipping normal fault, representing the (Fig. 8.3), as shown by a poorly-defined, inclined seismic eastward extension of the South Samana Bay and Septen- zone that strikes approximately east-west56,68. This inclined trional Faults, that is referred to as the 19° Latitude Fault; seismic zone is the northward continuation of the east-dip- and the north-dipping Escarpment Fault, responsible for ping Lesser Antilles subduction zone7. Focal mechanism active normal faulting and the graben form of the western

152 Puerto Rico and the Virgin Islands

Figure 8.2. Generalised structural blodcs of the northeastern Caribbean (based on numerous sources). Abbreviations: A=Antigua; An=Anguilla; Ba=Barbuda; KT=Kallinago Trough; N=Nevis; PR=Puerto Rico; S=Saba; SB=St. Barts; SM=St. Martin; VI=Virgin Islands.

Puerto Rico Trench. The 19° Latitude Fault separates a of the Mona Passage is a north-trending graben, formed by metamorphic or blueschist complex, called the North Slope extension between Puerto Rico and Hispaniola25,48. The complex herein, on the inner trench wall33'34'64 from the arc trace of the Mona Canyon is complicated in the southern massif proper, consisting of island arc rocks overlain by part of the Mona Passage by a series of south-moving shallow water limestones64. extensional allochthons, referred to as the Mona extensional In the Muertos Trench, seismic reflections, focal allochthon48. mechanisms and earthquake epicentres show northward un- Based on dredge haul and submersible obs ervations, derthrusting of the Venezuelan Basin nearly normal to the the shelf of Puerto Rico extended to the 19° Latitude Fault trench4,11,41-43. As indicated by seismic reflection and up until the middle Miocene2,5,23,57,61,64,70,83. At that time, GLORIA data, the amount of subduction decreases toward profound subsidence dropped the northern part of the arc the east, such that Muertos Trench subduction is thought to massif down to water depths in excess of 4 km. Such end south of St. Croix (the trench is pinned at this posi- subsidence is believed to have been associated with listric tion)20,31,53. The Muertos Trench accretionary complex is normal motion (Fig. 8.3) on a series of faults from the 19° thought to be composed mostly of material scraped off and Latitude Fault to the Puerto Rico Trench46,76, although other transferred from the underthrusting Venezuelan Basin. The models have been proposed invoking trench initiation2 and nature of the contact between the PRNVI Block and the tectonic erosion . Based on the mapped distribution of the Muertos accretionary complex is not understood. rollover, vertical displacement on the 19° Latitude Fault The St. Croix Block is separated from the northern seems to decrease toward the northeast and a pinning point Lesser Antilles island arc (Saba and Anguilla banks; Fig. northeast of Puerto Rico has been proposed46,76. 8.2) by a fault zone probably characterized by extension and, Palaeomagnetic studies of Oligocene to Pliocene lime- perhaps, oblique motion12,31. The PRNVI Block si sepa- stones on Puerto Rico indicate about 25° of anticlockwise rated from Hispaniola by the Mona Passage, a seaway rotation of the PRNVI Block in the late Miocene67, broadly connecting the Atlantic Ocean and the Caribbean Sea. The contemporaneous with the normal faulting and subsidence Mona Canyon (Figs 8.1, 8.2) underlying the northern part (19° Latitude Fault, Mona Canyon, Mona Allochthon,

153 DAVID K. LARUE

Anegada Passage, Virgin Islands Basin) alluded to above. Oligocene, than the northern Virgin Islands and St. Croix, The rotation history of other tectonic elements in the region where carbonate sedimentation probably commenced in the is not well constrained. Miocene. Puerto Rico Interpretation of neotectonics The middle to late Tertiary of Puerto Rico is usually A simple model which evokes rotation of the arc massif considered in terms of the North Coast and South Coast of the PRNVI Block seems to explain several important Basins (Figs 8.1, 8.4). Information about these basins has neotectonic features46,67. It is assumed that the Muertos been gained from outcrops in the foothills of the central Trench is apparently pinned (that is, displacement on the mountains of Puerto Rico, from shallow boreholes for water fault is essentially zero or it is in the centre of rotation of the (less than 925 m) and from five exploratory wells for petro- fault) south of St. Croix and, similarly, the 19° Latitude Fault leum (CPR 1-4 and Toa Baja #1; Fig. 8.1). All of the seems to be pinned northeast of Puerto Rico. These pinning petroleum wells were dry, although traces of methane were points help define an axis of rotation for the PRNVI Block found at Toa Baja #144. Additionally, seismic reflection data at about 18°N and 65°W76. Underthrusting in the Puerto are available locally for the shelf and slope of Puerto Rico. Rico Trench apparently increases toward the east and exten- The boundaries of the basins are not clear in all cases. The sional tectonism (for example, the 19° Latitude and Escarp- southern and northern boundaries of the North Coast and ment Faults) increases towards the west. Underthrusting in South Coast Basins, respectively, are sedimentary onlaps the Muertos Trench increases toward the west and exten- above older deformed rocks. The submerged northern edge sional tectonism (Virgin Islands Basin, Anegada Passage) of the North Coast Basin is clearly cut by the 19° Latitude increases towards the east. Finally, palaeomagnetic evi- and Escarpment Faults. The southern boundary of the South dence indicates that the PRNVI Block rotated 25° anticlock- Coast Basin is not defined by available data. wise in the late Miocene to early Pliocene. Timing of the The North Mona Basin occurs in the northern part of various tectonic events is no\ excellent, but rotation seems the Mona Passage (Fig. 8.1), is entirely submarine and is broadly synchronous with deformation in the Muertos defined on the basis of seismic reflection data57. The San Trough and the 19° Latitude Fault, and extension in the Juan Basin is the eastward equivalent of the North Coast Virgin Islands Basin. Therefore, it seems that observable Basin, but is separated from it by a structural high, the San neotectonic features are probably associated with this rota- Juan high. The San Juan Basin is poorly understood and tional event: considering the PRNVI Block to be roughly definition is based on limited seismic data. rectangular in plan, shortening occurred on the northeast and Stratigraphic units in the North Coast Basin are defined southwest coiners, and extension occurred on the northwest in Figure 8.559,71. The Upper Oligocene San Sebastian and southeast corners. The cause of the tectonic rotation of Formation lies unconformably over deformed Cretaceous the PRNVI Block may be due to the passage of Puerto Rico and Eocene rocks. It is comprised of terrestrial and shallow- and the Virgin Islands past the prong of the Bahamas (Fig. marine sandstones, mudstones and conglomerates derived 8.1), which is inhibiting movement of Hispaniola, and the from the erosion of older rocks, with limestones developed dominance of strike-slip tectonism between the North locally. The Upper Oligocene to Lower Miocene Lares American and Caribbean Plates. Puerto Rico and the Virgin Limestone conformably overlies the San Sebastian Forma- Islands may therefore be undergoing 'tectonic escape' relation and represents a carbonate platform facies, as do the tive to the Greater Antilles further west of the Bahamas overlying Lower to Middle Miocene Cibao Formation, Los prong. In addition to extension associated with block rota- Puertos Limestone and Aymamon Limestone. This late tion, studies of global plate motion seem to indicate a Oligocene to middle Miocene carbonate platform appar- regional component of extension between the Caribbean and ently extended all the way to the Puerto Rico Escarpment North American Plates in the vicinity of Puerto Rico32,76. (19° Latitude Fault), based on studies of dredge hauls and However, differentiating the two extensional components submersible observations. produced by block rotation and plate divergence is difficult. The Pliocene Quebradillas Limestone unconformably overlies the Aymamon Limestone and includes some rocks 61 PART H: MIDDLE TO LATE TERTIARY of bathyal facies. Moussa et al. attributed this change in STRATIGRAPHY depositional style to subsidence of the north slope of Puerto Rico to depths in excess of 4 km near the Puerto Rico From the Oligocene or Miocene to the Recent, Puerto Rico Escarpment. and the Virgin Islands were complex carbonate banks, with The South Coast Basin contains a similar sequence of sedimentation influenced by local tectonism. Time of bank strata (Fig. 8.5), although influence of tectonism during initiation was earlier in Puerto Rico, in the middle to late sedimentation is indicated by variations in stratigraphic

154 Puerto Rico and the Virgin Islands

Figure 8.3. Interpretive and generalized north-south cross-section through (A) the Mona Passage and (B) the Virgin Islands. No vertical exaggeration.

thicknesses and nature of basin fill24,59,78. The lowermost St. Croix unit is the Middle to Upper Oligocene Juana Diaz Forma- Middle to late Tertiary rocks on St. Croix are confined tion. It is comprised of: a basal clastic sedimentary member to the Kingshill Graben, located near the west-central part composed of material derived from the weathering of the of the island. Pelagic and hemipelagic carbonates of the ancestral mountains of Puerto Rico; an intermediate reefal Kingshill Limestone overlie blue pelagic and hemipelagic limestone unit similar to the Lares Limestone; and an upper carbonates of the Jealousy Formation27 (Fig. 8.5). The island-slope chalk member called the Angola Limestone. Lower to Middle Miocene Jealousy Formation was pre- Deposition of the lower unit occurred locally in a deepwater viously thought to be Oligocene. Deposition of the Jealousy basin and elsewhere in shallow water24,60,78. Unconfor- Formation and lower Kingshill Limestone occurred in deep- mably overlying the Juana Diaz Formation is the late Mio- water (around 600 m), whereas the upper Kingshill Lime- cene Ponce Limestone, which is comprised of shallow-water stone accumulated in about 200 m of water. The youngest carbonates. The South Coast Basin, complicated by normal unit is the Blessing Formation, which represents the culmi- faults of post-middle Miocene age, is more structurally nation of continued shoaling of the Kingshill Basin and is disrupted than the North Coast Basin, which underwent a comprised in part of reef facies27. flexural event in the post-middle Miocene with insignificant Summary of middle Tertiary tectonic evolution amounts of faulting. Deposition of Oligocene to Miocene rocks in the Puerto Northern Virgin Islands Rico and Virgin Islands region occurred primarily in a The Upper Miocene to Holocene, Rogues Bay Forma- quiescent carbonate bank setting. Late Oligocene tectomsm tion of Tortola is a thin (10-13 m), shallow-water carbonate is indicated by the variable nature of basin fill in the South unit which dips at 15-20° to the west30. The active carbonate Coast Basin of Puerto Rico, but not elsewhere. Late early platform of the northern Virgin Islands, which is comprised Miocene extension and graben formation probably occurred of hundreds of metres of limestones, was probably initiated in the St. Croix Block, forming the Kingshill Basin27. Car- in the late Miocene (Fig. 8.5). bonate bank sedimentation was disrupted in the late Mio- cene by subsidence on the north coast and extensional

155 DAVID K. LARUE

tectonism on the south coast of Puerto Rico. Bermeja Complex records the history of early growth of an island arc, from a basement composed of oceanic oust to PART 1H. EVOLUTION OF THE EASTERN arc volcanic materials. This early arc sequence was dis- GREATER ANTILLES VOLCANIC ISLAND ARC rupted for unknown reasons prior to the late Cretaceous during deformation of the serpentinite melange. Much of the rock exposed on Puerto Rico and the Virgin Upper Cretaceous formations (Fig. 8.6) in the South- Islands is of Cretaceous to Eocene age (to Oligocene in the west Block consist of limestones, volcanic rocks such as northern Virgin Islands) and records the evolution of the lava flows and shallow intrusives, and volcaniclastic Greater Antilles island arc. The volcanic arc was complex rocks1,14,15,54,81. Sedimentological studies of the limestones and the geology is perhaps best understood by subdividing the indicate that two major depositional facies are preserved, arc into quasi-homogenous geologic domains. shallow-water and slope/basinal. Slope and basinal facies Jurassic to early Tertiary rocks of Puerto Rico are occur in the Parguera Limestone, with isolated exposures in usually separated into three litho-tectonic blocks, the South- the Yauco Formation, whereas the Cotui Limestone consists west, Central and Northeast. Recent research by Larue et al. 47 of shallow-water (including rudist-dominated) facies. has defined four litho-tectonic blocks, dividing the Southwest Depositional facies or ages of volcaniclastic rocks have not Block into the Southwest and Southc entral Blocks. These been studied adequately. Therefore, it is not certain what blocks separate different structural and stratigraphic domains, controls the periodicity of cyclic volcaniclastic and lime- and are fault-bounded (Fig. 8.4). stone deposition. The geology of Culebra seems to be strongly similar to the Southcentral Block Northeast Block of Puerto Rico, although individual rock The Southcentral Block is separated from the Southwest units cannot be unequivocally correlated. In contrast, Vieques Block by the last exposure of serpentinite (Fig. 8.4) of the has been correlated with the Central Block of Puerto Rico. Bermeja Complex47. The nature of this boundary is The thick Cretaceous to Eocene sequence in the northern uncertain. Some workers have claimed that it is possible to Virgin Islands cannot be correlated exactly with rocks in Puerto map formations across the boundary54, although it is not Rico, necessitating the definition of another block, although a clear whether the formations are truly correlative. comparable history is indicated. However, the geology of St. Upper Cretaceous rocks of the Southcentral Block in- Croix cannot be tied with certainty to any other block in the 85 clude the Yauco, Lago Garzas, Maricao and related, domi- Puerto Rico-Virgin Islands area (although Whetten has nantly volcaniclastic, Formations 38"40 (Fig. 8.6). Field argued that keratophyre clasts in Cretaceous rocks on St. studies47 indicate deposition occurred in deep-water, pre- Croix were derived from the northern Virgin Islands). dominantly in the turbidite realm. No basement rocks are The Turtlehead Block occurs between the northern exposed in this region. Several Cretaceous to Eocene plu- Virgin Islands and the trench slope break of the Puerto Rico tons cut the basinal sequence. Limestone units of uncertain Trench. Seismic reflection profiles indicate that the early age and origin have been described. Tertiary strata of this block are possibly less deformed than on Central Block the northern Virgin Islands (Fig. 8.3). The Central Block is bound by the Great Northern and Southern Fault Zones (Fig. 8.4). The pre-Robles Group (Fig. Mesozoic geology 8.6) represents the oldest strata in central Puerto Rico and Southwest Block of Puerto Rico has been divided by the U.S. Geological Survey into four The oldest rocks in Puerto Rico occur in the Bermeja formations, A to D. The pre-Robles Group is comprised of Complex (Fig. 8.6), a serpentinite melange consisting of both shallow -water and probable deep-water deposits rep- dismembered ocean floor and island arc derivatives such as resenting facies equivalents and temporal changes in water volcanic and volcaniclastic rocks54'69'80"82. Cherts as old as 55 depth. Shallow-water deposits are indicated by intercalated Kimmeridgian have been described from this block . Am- massive limestones bearing rudists, such as the Aguas phibolites are thought to represent the original basement of Buenas Limestone (base of Formation D) and a limestone the Bermeja Complex, which have a K-Ar age of about 125 69 lens in Formation B. Overlying the pre-Robles Group are the Ma . Such young ages are thought to represent the age of Rio Orocovis and Robles Groups, which are apparently recrystallization associated with metamorphism. stratigraphically equivalent. The contact between the pre- Lower Cretaceous igneous rocks in the Bermeja Complex Robles and Robles Groups is uncertain in nature and age. seem to have an oceanic island arc affinity, whereas The Rio Orocovis and Robles Groups are basinal deposits, amphibolites and associated cherts may represent rocks of representing a deepening-upwards sequence35. However, a 69 oceanic crustal origin . It has therefore been argued that the limestone lens occurs in the Avispa Formation near the top of the Rio Orocovis Group, indicating shoaling following

156 Puerto Rico and the Virgin Islands

Figure 8.4. Geologic map of Puerto Rico and the Virgin Islands (redrawn after Garrison et al 26).

initial subsidence. Upper Cretaceous subaerial and littoral Northern Virgin Islands volcaniclastic deposits (including welded tuffs) of the A homoclinal sequence of Cretaceous to Oligocene Pozas Formation cap the Cretaceous sequence3'8. rocks, with a structural thickness of over 20 km, is exposed The Central Block also possesses two large plutons, in the northern Virgin Islands (Fig. 8.4). Donnelly16,17 the Utuado and San Lorenzo Batholiths, as well as a demonstrated the island arc affinity of this sequence and number of smaller intrusive bodies. The age of plutonism argued that it represents a cross-section through an evolving ranges from Cretaceous to late Eocene13. oceanic island arc. The lowermost structural unit exposed in Northeast Block outcrop is the poorly-dated Water Island Formation, possi- In northeast Puerto Rico, the oldest rocks are deep- bly of early Cretaceous (Albian49) antiquity, greater than 5 water basinal deposits of the Daguao Formation, km in structural thickness, and comprised of deep-water Figuera Lava, Fajardo Formation and Tabonuco lava flows with associated breccias, tuffs and minor radio- Formation (Fig. 8.6). Shallow-water limestone units are larites. The minor occurrences of tuff may reflect a hyalo- present higher in the sequence, including the Limestone clastic origin, similar to hyaloclastites found on seamounts. Lens in the El Ocho Formation, the Campanian or A "soft, sticky clay", including the minerals chlorite younger Rio de la Plata Limestone, and the Maastnchtian and smectite, has been penetrated by drilling the base of the La Muda and Trujillo Alto Limestones, indicating a Water Island Formation at a depth of 725 m on southeast St. shoaling sequence. John18,29. This deformation zone may suggest that the The Northeast Block contains a similar sequence of Water Island Formation is thrust to the south over less rocks to those of the Central Block, including Lower metamorphosed sediments (Fig. 8.3B). and Upper Cretaceous volcanic and volcaniclastic rocks 84, The overlying Virgin Islands Group, of possible and limestones (Fig. 8.6). However, unlike the Central Cenomanian17,30 or Turonian to early Santonian age, is Block, the Northeast Block does not contain major structurally about 8 km thick, and consists of epiclastic and plutons, although it has been cut by east-west trending pyroclastic strata, including breccias and volcaniclastic tur- basaltic dyke swarms, apparently of early Tertiary age. bidites, with local limestone members. Resedimented shal- Puerto Rican Virgin Islands low-water sediments are indicated by fossils and rounded Vieques and Culebra (Figs. 8.1, 8.4) domin ate the conglomerate clasts. Moreover, clasts of the underlying Puerto Rican Virgin Islands, but little is known about Water Island Formation are abundant in the basal part of the their geology36. Upper Cretaceous volcaniclastic rocks Virgin Island Group strata. The Eocene Tortola Formation are present on both islands and Vieques contains a large consists mostly of deep-water volcaniclastic strata, but also granite pluton. Vieques therefore has some affinity with contains a possibly resedimented(?), shallow-water lime- 30 the Central Block of Puerto Rico, whereas Culebras bears stone facies . Subaerial volc anic deposits, termed the some resemblance to the Northeast Block9. Necker Formation, of Oligocene(?) age, are intruded by the

157 DAVID K. LARUE

Figure 8.5. Cenozoic stratigraphy of Puerto Rico and the Virgin Islands (based on many sources).

Oligocene Virgin Islands Batholith and overlain by Upper apparently not undergone the same deformation history, Miocene to Holocene limestones of the Rogues Bay Forma- appearing less deformed. The Turtlehead well bottomed in tion30,51. igneous (or volcaniclastic) rocks of uncertain age after pene- The northern Virgin islands therefore represent a rela- trating Tertiary sediments. This may indicate the presence of tively complete island arc sequence, although the actual another tectonic block, termed here the Turtlehead Block. oceanic arc basement is not exposed. It should be noted that St. Croix 66 Rankin argued against a simple homoclinal stack and for a Cretaceous rocks on St. Croix are similar to those more complex megafold based on regional studies of observed on Puerto Rico and the Virgin Islands, consisting of sedimentary facing-directions. Upper Cretaceous (Cenomanian to Maastrichtian), deep- Turtlehead water, volcaniclastic sediments, intruded locally by Upper Seismic reflection profiles and correlation with Turtle- Cretaceous (Maastrichtian) plutons of mafic to intermediate head #1, drilled by Mobil in the 1980s (Fig. 8.1), indicates composition72-74,85. Formations mapped on St. Croix are that strata to the north of the northern Virgin Islands have principally of submarine fan facies, such as massive sand-

158 Puerto Rico and the Virgin Islands

Figure 8.6. Mesozoic stratigraphy of Puerto Rico and the Virgin Islands (based on many sources).

stones and thin-bedded turbidites. An unusual feature of the tions. Sedimentary rocks in the north slope complex are from submarine fan facies rocks is the common occurrence of Eocene to Holocene in age. The metamorphic rocks probably black cherts, not clearly of pelagic origin. The major oddity of came from arc-related protoliths. St. Croix relative to the rest of the Virgin Islands and Puerto Early Tertiary stratigraphy of Puerto Rico Rico is the intense deformation of the rocks, including possible The early Tertiary stratigraphy of Puerto Rico is not clearly thrust faults, and the widespread development of a penetrative related to the blocks previously defined or to early Tertiary rocks slaty cleavage (see below). Whetten85 argued that keratophyric that occur between the blocks. The Eocene Belt of Puerto Rico fragments in Cretaceous sandstones on St. Croix may have been separates the Southcentral and Central Blocks, and is structurally derived from the northern Virgin Islands. If so, this provides associated with the Great Southern Fault Zone. Elsewhere, early the only direct geologic link, other than the general Tertiary rocks occur fairly sporadically or else are preserved in composition of the Upper Cretaceous arc complex, with any grabens (Fig. 8.4). Early Tertiary rocks also have a scattered other island of the eastern Greater Antilles. distribution between the Central and Northeastern Blocks (near North Slope complex the Great Northern Fault Zone), though a belt of rocks per se is As mentioned in the neotectonics section, a metamorphic not well-defined as the Eocene belt. Early Tertiary rocks include complex is recognised between the 19° Latitude Fault and the volcanic flows of basic to intermediate composition, Puerto Rico Trench, based on analysis of dredge hauls, volcaniclastic turbidites, shallow - and deep-water limestones, studies from submersibles and regional geologic and intru sions. An interesting feature of early Tertiary rocks in considerations. Perfit et al 64 concluded that the metamor- north central Puerto Rico is the local occurrence of plutonic rock phic rocks may be continuous with rocks of the Samana Pen- fragments and detrital quartz grains, indicating that the insula on Hispaniola33,34 and represent moderate P-T condi- plutons of the island were being uplifted and eroded. Shal-

159 DAVID K. LARUE

low-water limestone facies are found in northcentral and have been mapped on Puerto Rico, the Great Northern and southcentral Puerto Rico, suggesting that the Central Block Southern Fault Zones (Fig. 8.4). Deformation associated with may have been a topographic high in the Eocene. Locally, these fault zones seems to be primarily strike-slip. However, fragments of Cretaceous limestones are present in Eocene evidence of shortening is also widespread, as indicated by strata of south central Puerto Rico. Other early Tertiary folded and thrust-faulted rocks, especially in southcentral facies seem to be of a deep-water origin. Puerto Rico21,22,28 (Fig. 8.4). Uplift, as indicated by Early Tertiary stratigraphy of the northern Virgin Islands metamorphic facies below the unconformity, was of the order Two Paleogene stratigraphic units are recognised in the of 2 or more kilometres. Together the two effects seem to indicate 10,22 northern Virgin islands, the Tortola and Necker Formations a dominantly transpressional deformation environment . (Fig. 8.5). The Eocene Tortola Formation is over 8 km in Also associated with this deformation was anticlockwise block structural thickness, and is comprised of andesitic and rotation of at least 20°, based on studies of palaeomagnetism in 79 augite-andesitic tuffs, breccias and volcanic sandstones de- Eocene rocks . posited in a submarine setting. An intercalated shallow-water Northern Virgin Islands limestone lens may be resedimented. This unit is overlain by Deformation occurred later in the northern Virgin Islands the Oligocene(?) Necker Formation, which is approximately than in Puerto Rico, where plutons and intrusives as young as 13,37 2 km in structural thickness, and is comprised of tuffs and Upper Oligocene occur which are overlain by a poorlv- 30 breccias deposited under subaerial conditions. Longshore51 developed, Miocene to Pliocene overlap sequence . The argued that the Oligocene13 Virgin Islands Batholith predates northern Virgin Islands consist of a great homoclinal sequence or is the same age as the Necker Formation. locally cut by faults (see above). Outcrop-scale folds are very Early Tertiary of St. Croix rare, although the rocks are locally foliated. It is not clear whether No Paleogene rocks are present on St. Croix. Older this foliation is related to the regional tectonic event or to the publications suggested that the Jealousy Formation may be as intrusion of the Virgin Islands Batholith. On St. Thomas, the old as Eocene. However, recent studies indicate that it is foliation strikes approximately east-west and dips essentially Miocene. The oldest pelagic carbonates on St. Croix are vertically. known from a mudball (Fig. 8.5) which contains foraminif- Based on seismic reflection data and information from the ers of early Eocene to early Miocene age50. Turtlehead well (Fig. 8.1), it can be concluded that Eocene Plutons rocks north (=offshore) of Tortola in the Turtlehead Basin are not Plutons of varying size and composition occur in Puerto significantly deformed. This suggests that the intensity of the Rico and the Virgin Islands. Plutons in Puerto Rico are Eocene-Oligocene deformational event decreases toward the largest in the Central Block, are granodioritic, dioritic and north. A similar situation appears to apply north of Puerto 45 quartz dioritic in composition, and are clearly of calc -alka- Rico , where undeformed Eocene strata are apparently present line affinity62. The Utuado and San Lorenzo Batholiths are late to the north of folded and faulted Eocene rocks. This suggests Cretaceous in age (80-60 Ma) 13, whereas other plutons on that a deformation front exists within the arc crust, separating Puerto Rico range from 130-37 Ma. The Virgin Islands less deformed rocks in the north from more deformed rocks in Batholith is younger, 36-33 Ma, but locally is as young as the south. 24.3 Ma13,37. Intrusive rocks, including dykes, sills and larger bodies referred to as plutons, are present on St. Croix, St.Croix ranging in age from 71.8-66.1 Ma74. It is not clear whether St. Croix experienced this Eo- cene-Oligocene deformational event. Speed et al74 pre- Eocene to Oligocene deformational events viously proposed that deformation occurred in the Puerto Rico Cretaceous, based on the observation of non-deformed Cre- Localized intense deformation occurred between depo- taceous intrusions cutting foliated rocks. However, foliated sition of Eocene and Oligocene strata, and is indicated by a intrusive bodies have now been discovered, indicating at 19 profound angular unconformity. Timing of this deformation least some of the plutons are pre-deformational . Moreover, is based on the youngest age of the deformed sequence dredge samples from the northern slope of St. Croix reveal (middle to late Eocene) and the oldest age of the overlap volcaniclastic rocks similar to those on St. Croix, but with sequence (middle to late Oligocene), indicating that defor- whole-rock K-Ar ages as young as 33.5-28.3 Ma 6 6 mation occurred in the late Eocene to middle Oligocene. (Oligocene) . Bouysse et al. interpreted the younger age to be Style of deformation is variable, and not simply defined by the age of seritization accompanying development of end-member deformation types such as fold and thrust belt, schistosity. Thus, it appears that St. Croix may have been or zone of strike-slip, tectonism. Two principal fault zones affected by Oligocene tectonism. Style of deformation on St. Croix is unique compared

160 Puerto Rico and the Virgin Islands

to elsewhere in the eastern Greater Antilles, consisting of SUMMARY OF NINE PHASES OF DEVELOPMENT OF PUERTO RICO AND tightly-folded sandstones and mudstones, typically exhibit- THE VIRGIN ISLANDS ing a well-developed axial planar foliation, locally cut by 75 faults. Speed and Joyce recently reinterpreted the style of (1) Oceanic crustal development in the late Jurassic, as deformation using the fold and thrust belt model, and argued indicated by amphibolites in the Bermeja Complex69. that a number of discrete thrust sheets jean be recognised, a 19 (2) Early arc build-up in the early Cretaceous, as indicated view supported by Draper and Bartel . by arc volcanics in the Bermeja Complex, cutting the oceanic crustal rocks, and volcanism and associated Summary of Cretaceous and early Tertiary deformation events sedimentation in the Central Block of Puerto Rico (pre- A total of three deformation events are currently recog- Robles Group) and the northern Virgin Islands (Water nised in the Cretaceous to early Tertiary of Puerto Rico and Island Formation). the Virgin Islands. These deformation events are not univer - (3) Early arc disruption, as indicated by the Bermeja Com sally correlative between islands. The oldest event, that plex serpentinite melange, which contains blocks of formed the serpentinite melange of the Bermeja Complex, both arc and oceanic crustal rocks. This early arc dis is recognised in southwest Puerto Rico. This serpentinite ruption phase is manifest elsewhere in Puerto Rico and melange is comprised of rocks with both arc and oceanic in the northern Virgin Islands by an unconformity of crustal affinities, and is overlain unconformably by the Albian age. Turonian Paguera Limestone. Deformation may therefore (4) Renewed arc build-up in the late Cretaceous (late Al- have occurred in the Aptian to Cenomanian. A key point is bian(?) to Cenomanian, continuing to the Maas- that blocks in the melange consist of oceanic crustal rocks trichtian), as indicated by a thick volcanic pile in Puerto intruded by arc volcanic rocks, indicating the deformation Rico, the northern Virgin Islands and St. Croix. of an island arc already built atop oceanic crust and is (5) A localised end-Cretaceous (mostly Maastrichtian) arc therefore not an arc-oceanic crust accretion event. The age disruption event, recognised only in St. Croix. of the melange formation in the Southwest Block apparently (6) Renewed arc build-up in the early Tertiary, associated corresponds to the age of the unconformity at the top of the with arc dissection as indicated by plutonic clasts, de- Water Island Formation and the pre-Robles Group (Fig. trital quartz and fragments of Cretaceous limestone in 8.6). early Tertiary rocks. This is an odd period of arc growth In St. Croix, a late Cretaceous deformation event is and uplift, probably transitional to the next phase of recognised as responsible for the intense deformation of evolution. Cenomanian to Maastrichtain volcaniclastic strata. Speed (7) Cessation of volcanism, uplift of several kilometres, and Joyce argued that deformation occurred in an accre- deformation and rotation of the arc massif in the late tionary prism setting, with plutonism occurring after trans- Eocene to middle Oligocene in Puerto Rico, and in the port of the accretionary prism over the arc massif. However, middle to late Oligocene in the northern Virgin Islands. an intra-arc basin deformational site seems more likely, This deformation is possibly associated with subduc especially based on the observation of Draper and Bartel19 tion of buoyant oceanic crust or related to a change in of deformed plutons, indicating that sedimentation occurred plate motion associated indirectly with the Cuban col- in an island arc environment, not a forearc. This deformation lision to the west. event or a correlative unconformity has not been recognised (8) Carbonate platform development on Puerto Rico (late on Puerto Rico or the northern Virgin Islands. Oligocene to Miocene) and the northern Virgin Islands The early Tertiary event recognised in Puerto Rico and (late Miocene?), with sedimentation in abasinal setting the northern Virgin Islands is clearly the most widespread in St. Croix (early Miocene). of the three deformations recognised. Origin of the defor- (9) A deformational event from middle Miocene to the mation is uncertain, but most recent studies indicate a tran- Recent, associated with profound, local, tectonic subsi spressional setting21,22,28, possibly associated with dence, local extensional and contractional tectonism, subduction of buoyant seafloor46. Deformation apparently and accompanied by anticlockwise block rotation of decreases toward the north of Puerto Rico and the northern Puerto Rico and the northern Virgin Islands by 25°. Virgin Islands, supporting the intra-arc transpressional theory. The importance of this orogenic event in St. Croix is ACKNOWLEDGMENTS—This chapter represents a summary of work performed at the University of Puerto Rico from 1986-1991. Discussions with unclear. colleagues there were always fruitful, especially James Joyce, Hans Schellekens and Alan Smith. Support for these studies came from the Puerto Rico Power Authority and included two grants from the National Science Foundation. Reviews by Gren Draper and an anonymous reviewer

161 DAVID K.LARUE

improved the typescript. ogy of some metamorphic, igneous and hydrothermal events in Puerto Rico and the Virgin Islands. U.S. Geological Survey Journal of Research, 5, 689-703. REFERENCES 14Curet, A.F. 1981. The geology of a Cretaceous-Tertiary 1 volcano-sedimentary sequence in the''Mayaguez and Almy, C. C., Jr. 1965. Parguera Limestone, Upper Creta- Rosario Quadrangles in west-central Puerto Rico. Un- ceous Mayaguez Group, southwest Puerto Rico. Un- published Ph.D. thesis, University of California at published Ph.D. thesis, Rice University, Houston, 189pp. 2 Santa Barbara Alonso, R.M., Krieg, E.A. & Meyeitoff, A. A. 1987. Post- 15Curet, A.F. 1986. Geologic map of the Mayaguez amd early Pliocene age of the Puerto Rico Trench: in Duque- Rosario Quadrangles, Puerto Rico. U.S. Geological Caro, H. (ed.), Transactions of the Tenth Caribbean Survey, Miscellaneous Geologic Investigations, Map 1- Geological Conference, Cartagena, Columbia, 14th- 1657. 20th August, 1983, 82-103. 16 3 Donnelly, T.W. 1964. Evolution of eastern Greater Antilles Berryhill, H.L., Jr. 1965. Geology of the Ciales Quadrangle, island arc. American Association of Petroleum Puerto Rico. U.S. Geological Survey Bulletin, 1184, Geologists Bulletin, 48, 680-696. 84pp. 17 4 Donnelly, T.W. 1966. Geology of St. Thomas and St. John, Biju-Duval, B., Bizon, G., Mascle, A. & Muller, C. 1982. U.S. Virgin Islands: in Hess, H.H. (ed.), Caribbean Active margin processes: field observation in southern Geological Investigations. Geological Society of America Hispaniola: in Watkins, J.S. & Drake, C.L. (eds), Studies Memoir, 98, 85-176. in Continental Margin Geology. American Association of 18Donnelly, T.W. 1968. Field guide to the geology of St. Petroleum Geologists Memoir, 34, 325-344. 5 John and St. Thomas. Field Trip Guidebook, Fifth Birch, F.S. 1986. Isostatic, thermal and flexural models of the Caribbean Geological Conference, St. Thomas, U.S. subsidence of the north coast of Puerto Rico. Geology, Virgin Islands, 1-5 July, 1968,1-26. 14, 427-429. 19 6 Draper, G. & Bartel, J. M. 1990. Development of cleavage in Bouysse, P., Andreieff, P., Richard, M, Baubron, J., Mascle, the Cretaceous rocks of St. Croix, U.S. V.I., and some A., Maury, R. &. Westercamp, D. 1985. Aves Swell and tectonic implications. Geological Society of America northern Lesser Antilles Ridge: rock dredging results Abstracts with Programs, 22, A337. from Arcante 3 cruise: in Mascle, A. (ed.), Geody- 20EEZ SCAN 85 Scientific Staff. 1987. Atlas of the U.S. namiquesdes Caribes, Symposium, Paris, 5-8Fevrier. Exclusive Economic Zone, eastern Caribbean. U.S. Editions Technip, Paris, 65-75. 7 Geological Survey, Miscellaneous Geologic Investiga- Bowin, C. 1972. Puerto Rico Trench negative gravity tions, I-1864B, 58 pp. anomaly belt. Geological Society of America Memoir, 21Erikson, J.P. 1988. Structural study of the Paleogene 132, 339-350. 8 deformation in and around the Southern Puerto Rico Briggs, R.P. 1971. Geologic map of the Orocovis Quadrangle, Fault Zone of south central Puerto Rico. Unpublished Puerto Rico. U.S. Geological Survey, Miscellaneous Ph.D. thesis, Stanford University, 79 pp. Geologic Investigations, Map 1-615. 22 9 Erikson, J.P., Pindell, J.L. & Larue, D.K. 1989. Mid Briggs, R.P. & Akers, J.P. 1974. Hydrogeologic map of Eocene-early Oligocene sinistral transcurrent faulting in Puerto Rico and adjacent islands. U.S. Geological Survey Puerto Rico associated with formation of the northern Hydrologic Investigations Atlas, HA-197. Caribbean Plate Boundary Zone. Journal of Geology, 10Briggs, R.P. & Pease, M.H. 1961. Compressional graben and 98, 365-384. horst structures in east-central Puerto Rico. U.S. 23Fox, P.J. & Heezen, B.C. 1975. Geology of the Caribbean Geological Survey Short Papers, B365-B366. 11 crust: in Nairn, A.E.M. & Stehle, F.G. (eds), The Ocean Byme, D.B., Suarez, G. & McCann, W.R. 1985. Muertos Basins and Margins. 3. The Gulf of Mexico and the Trough subduction-microplate tectonics in the northern Caribbean. Plenum, New York, 421-466. Caribbean. Nature, 317, 420-421. 24 12 Frost, S.H., Harbour, J.L., Beach, D.K., Realini, M.J. & Case, I.E. 1975. Geophysical studies in the Caribbean Harris, P.M. 1983. Oligocene reef tract development, Sea: in Nairn, A.E.M. & Stehli, F.G. (eds), The Ocean southwestern Puerto Rico. Sedimentia, 9, 141 pp. Basins and Margins. 3. The Gulf of Mexico and the Rosenstiel School of Marine and Atmospheric Science, Caribbean. Plenum, New York, 107-180. 13 Miami. Cox, D.F., Marvin, F.R., M'Gonigle, J.W., Mclntyre, 25Gardner, W.D., Glover, L.K. & Hollister, C.D. 1980. D.H. & Rogers, C. 1977. Potassium-argon geochronol- Canyons off northwest Puerto Rico: studies of their origin and maintenance with the nuclear research sub-

162 Puerto Rico and the Virgin Islands

marine NR-1. Marine Geology, 37, 41-70. Map I-1556. 39 26Ganison, L.E., Martin, R.G., Berryhill, H.L., Buell, Krushensky, R.D. & Monroe, W.H. 1978a. Geologic map of M.W., Ensminger, H.R. & Peny, R.K. 1972. Prelimi- the Penuelas and PuntaCuchara Quadrangles, Puerto Rico. nary tectonic map of the eastern Greater Antilles region. U.S. Geological Survey, Miscellaneous Geologic U.S. Geological Survey, Miscellaneous Geological In- Investigations, Map I-1042. 40 vestigations, Map 1-732. Krushensky, RD. & Monroe, W.H. 1978b. Geologic map of 27Gill, I.P., Hubbard, D.K., McLaughlin, P. & Moore, C. H. the Yauco and Punta Verraco Quadrangles, Puerto Rico. 1989. Sedimentological and tectonic evolution of Ter - U.S. Geological Survey, Miscellaneous Geologic tiary St. Croix: in Hubbard, D.K. (ed.), Terrestrial and Investigations, Map I-1147. Marine Geology of St. Croix, U.S. Virgin Islands. West 41Ladd, J.W., Shih, T. & Tsai, C.J. 1981. Cenozoic tectonics of Indies Laboratory, St. Croix, Special Publication, 8, central Hispaniola and adjacent Caribbean Sea. 49-72. American Association of Petroleum Geologists Bulletin, 28Glover, L. ffl. 1971. Geology of the Coamo area, Puerto 65,466-489. Rico, and its relation to the volcanic arc-trench associa- 42Ladd, J.W. & Watkins, J.S. 1978. Active margin structures tion. U.S. Geological Survey, Professional Paper, 636, within the north slope of the Muertos Trench. Geologie 102pp. en Mijnbouw, 57, 255-260. 29Hekinian, R. 1971. Petrological and geochemical study of 43Ladd, J.W, Worzel, J.L. & Watkins, J.S. 1977. Multifold spilites and associated rocks from St. John, U.S. Virgin seismic reflection records from the northern Venezuela Islands. Geological Society of America Bulletin, 82, Basin and the north slope of Muertos Trench: in Tal- 659-682. wani, M. & Pitman, W.C. (eds), Island Arcs, Deep-Sea 30Helsey, C.E. 1960. Geology of the British Virgin Islands. Trenches and Back-Arc Basins. American Geophysical Unpublished Ph.D. thesis, Princeton University. Union, Washington, D.C, 41-56. 31Houlgatte, E. 1983. Etude d'une partie de la frontiere 44Larue, D.K. 1990. Toa Baja drilling project, Puerto Rico. nord-est de las Plaque Caraibe. Unpublished thesis, EOS, 71, 233-234. Universite de Bretagne Occidental, 3° cycle, 69 pp. 45Larue, D.K. & Berrong, B. 1991. Cross section through the 32Jordan, T.H. 1975. The present day motions of the Carib- Toa Baja drillsite: evidence for northw ard change in late bean Plate. Journal of Geophysical Research, 80, 4433- Eocene deformation intensity. Geophysical Research 4440. Letters, 18, 561-564. 33 Joyce, J. 1982. The lithology and structure of the eclogite 46Larue, D. K., Joyce, J. & Ryan, H. F. 1990. Neotectonics of and glaucophane-bearing rocks on the Samana Penin- the Puerto Rico Trench: extensional tectonism and sula, Dominican Republic: in Snow, W., Gil, N., Llinas, forearc subsidence: in Larue, D.K. & Draper, G. (eds), R., Rodriguez-Torres, R., Seaward, M. & Tavares, I. Transactions of the Twelth Caribbean Geological Con- (eds), Transactions of the Ninth Caribbean Geological ference, St. Croix, U.S. Virgin Islands, 7th-IIthAugust, Conference, Santo Domingo, Dominican Republic, 1989,231-247. Miami Geological Society, Florida. 16th-20th August, 1980, 417-422. 47Larue, D.K., Pierce, P. & Erikson, J. 1991. Cretaceous 34 Joyce, J. 1985. High pressure-low temperature metamor- intra-arc summit basin on Puerto Rico: in Gillezeau, K. A. phism and the tectonic evolution of the Samana Penin- (ed.), Transactions of the Second Geological Conference sula, Dominican Republic, Greater Antilles. of the Geological Society of Trinidad and Tobago, Unpublished Ph.D. thesis, Northwestern University, Port-of-Spain, Trinidad, April 3-8, 1990, 184-190. 270pp. 48Larue, D.K. & Ryan, H.F. 1990. Extensional tectonism in 35 Kaczor, L. & Rogers, J. 1990. The Cretaceous Aguas the Mona Passage, Puerto Rico and Hispaniola: a pre- Buenas and Rio Maton Limestones of southern Puerto liminary study: in Larue, D.K. & Draper, G. (eds), Rico. Journal of South American Earth Sciences, 3, 1-8. 36 Transactions of the Twelth Caribbean Geological Con- Kemp, J.F. & Meyerhoff, H.A. 1926. Scientific survey of ference, St. Croix, U.S. Virgin Islands, 7th-11th August, Porto Rico and the Virgin Islands, volume 4, parts I and 1989,223-230. Miami Geological Society, Florida. II. New York Academy of Sciences, New York, 219 pp. 49Lewis, J.F. & Draper, G. (with Bourdon, C., Bowin, C, Kesler, S.E. & Sutter, J.F. 1979. Compositional Mattson, P., Maurrasse, F., Nagle, F. & Pardo, G.). evolution of intrusive rocks in the eastern Greater 1990. Geology and tectonic evolution of the northern Antilles island arc. Geology, 1, 197-200. 38 Caribbean margin: in Dengo, G. & Case, J.E. (eds), The Krushensky, R.D. & Curet, A.F. 1984. Geologic map of geology of North America. Volume H. The Caribbean the Monte Guilarte Quadrangle, Puerto Rico. U.S. Geo- region, 77 - 140. Geological Society of America, Boul- logical Survey, Miscellaneous Geologic Investigations,

163 DAVID K. LARUE

der. Earth, 3, 17-109. 64 50Lidz, B.H. 1984. Oldest (early Tertiary) subsurface car - Perfit, M.R., Heezen, B.C., Rawson, M. & Donnelly, T.W. bonate rocks of St. Croix, U.S. V.I., revealed in tur- 1980. Chemistry, origin, and tectonic significance of bidite-mudball. Journal of Foraminiferal Research, 14, metamorphic rocks from the Puerto Rico Trench. Ma- 213-227. rine Geology, 34,125-156. 51Longshore, J.D. 1966. Chemical and mineralogical vari- 65Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of ations in the Virgin Islands batholith and its associated the Caribbean region; a plate-tectonic perspective: in wall rocks. Unpublished Ph.D. thesis, Rice University, Dengo, G. & Case, J.E. (eds), The geology of North 94pp. America. Volume H. The Caribbean region, 405-432. 52Mann, P. & Burke, K. 1984. Neotectonics of the Carib- Geological Society of America, Boulder. bean. Review of Geophysics and Space Science,22, 66Rankin, D. 1984. Geology of the U.S. Virgin Islands, a 309-362. progress report. U.S. Geological Survey Open File Re- 53Masson, D.G. & Scanlon, K.M. 1991. The neotectonic port, 84-762, 83-96. setting of Puerto Rico. Geological Society of America 67Reid, J., Plumley, P. & Schellekens, J. 1991. Paleomag- Bulletin, 103,144-154. netic evidence for late Miocene counterclockwise rota- 54Mattson, P.H. 1960. Geology of the Mayaguez area, tion of north coast carbonate sequence, Puerto Rico. Puerto Rico. Geological Society of America Bulletin, Geophysical Research Letters, 18, 565-568. 71, 319-362. 68Schell, B.A. & Tarr, A.C. 1978. Plate tectonics of the 55Mattson, P.H. & Pessagno, E.A., Jr. 1979. Jurassic and northeastern Caribbean Sea region. Geologie en early Cretaceous radiolarians in Puerto Rican ophiolite- Mijnbouw, 51, 319-324. tectonic implications. Geology, 7,440-444. 69Schellekens, J.H., Montgomery, H., Joyce, J. & Smith, 56McCann, W.R. & Sykes, L.R. 1984. Subduction of aseis- A.L. 1990. Late Jurassic to late Cretaceous develop- mic ridges beneath the Caribbean Plate: implications ment of island arc crust in southwestern Puerto Rico: in for the tectonics and seismic potential of the north Larue, D.K. & Draper, G. (eds), Transactions of the eastern Caribbean. Journal of Geophysical Research, Twelth Caribbean Geological Conference, St. Croix, 89, 4493-4519. U.S. Virgin Islands, 7th-llth August, 1989, 268-281. 57Meyerhoff, A.A., Krieg, E.W., Cloos, J.D. & Taner, I. Miami Geological Society, Florida. 1983. Petroleum possibilities of Puerto Rico. Oil and 70Schneidermann, N., Beckmann, J.C. & Heezen, B.C. Gas Journal, 81 (no. 51), 113-120. 1972. Shallow water carbonates from the Puerto Rico 58Molnar, P. & Sykes, L.R. 1969. Tectonics of the Carib- Trench region: in Petzall, C. (ed.), Transactions of the bean and Middle America regions from focal mecha- Sixth Caribbean Geological Conference, Margarita nisms and seismicity. Geological Society of America Island, Venezuela, 6th-14thJuly, 1971, 423-425. Bulletin, 59, 801-854. 71Seiglie, G.A. & Moussa, M.T. 1984. Late Oligocene-Plio- 59 Monroe, W.H. 1980. Geology of the middle Tertiary cene transgressive-regressive cycles of sedimentation in formations of Puerto Rico. U.S. Geological Survey northwestern Puerto Rico: in Schlee, J. S. (ed.), Inter- Professional Paper, 954, 93 pp. regional Unconformities and Hydrocarbon Accumula- 60 Moussa, M.T. 1977. Bioclastic sediment gravity flow and tion. American Association of Petroleum Geologists submarine sliding in the Juana Diaz Formation, south- Memoir, 36, 89-96. western Puerto Rico. Journal of Sedimentary Petrol- 72Speed, R.C. 1974. Depositional realm and deformation of ogy, 47, 593-599. Cretaceous rocks, East End, St. Croix; in Multer, H.G. 61 Moussa, M.T., Seiglie, G.A., Meyeihoff, A.A. & Taner, & Gerhard, L.C. (eds), Guidebook to the Geology and I. 1987. The Quebradillas Limestone (Miocene-Plio- Ecology of some Marine and Terrestrial Environments, cene), northern Puerto Rico, and tectonics of the north- St. Croix, U.S. Virgin Islands. West Indies Laboratory, eastern Caribbean margin. Geological Society of St. Croix, Special Publication 5, 189-200. America Bulletin, 99 ,427-439. 73Speed, R.C. 1989. Tectonic evolution of St. Croix: impli- 62 Nelson, A.E. 1968. Intrusive rocks of north-central Puerto cations for tectonics of the northeastern Caribbean: in Rico. U.S. Geological Survey Professional Paper, Hubbard, D.K. (ed.), Terrestrial and Marine Geology of 600B, B16-B20. St. Croix, U.S. Virgin Islands. West Indies Laboratory, 63 Officer, C.B., Ewing, J.I., Hennion, D.G., Haikrider, D.G. St. Croix, Special Publication 8, 9-22. & Miller, D.E. 1959. Geophysical investigations of the 74Speed, R.C., Gerhard, L.C. & McKee, E.H. 1979. Ages of eastern Caribbean: Venezuela Basin, Antilles island deposition, deformation and intrusion of Cretaceous arc, and Puerto Rico Trench. Physical and Chemical rocks, eastern St. Croix, Virgin Islands. Geological

164 Puerto Rico and the Virgin Islands

Society of America Bulletin, 90, 629-632. Survey Bulletin, 1537-A, A73-A83. 75Speed, R.C. & Joyce, J. 1989. Depositional and structural 81 Volckmann, R.P. 1984a. Geologic map of the Puerto Real evolution of Cretaceous strata, St. Croix: in Quadrangle, southwest Puerto Rico. U.S. Geological Hubbard, D.K. (ed.), Terrestrial and Marine Survey, Miscellaneous Geologic Investigations, Map Geology of St. Croix, U.S. Virgin Islands. West Indies 1-559. Laboratory, St. Croix, Special Publication 8, 23-36. 82Volckmann, R.P. 1984b. Geologic map of the San Ger- 76Speed, R.C. & Larue, D.K. 1991. Extension and transten- man Quadrangle, southwest Puerto Rico. U.S. Geologi- sion in the plate boundary zone of the northeastern cal Survey, Miscellaneous Geologic Investigations, Caribbean. Geophysical Research Letters, 18,573- Map 1-558. 576. Talwani, M., Sutton, G.H. & Woreel, J.L. 1959. A 83Weaver, J., Smith, A. & Seiglie, G. 1975. Geology and crustal section across the Puerto Rico Trench. Journal tectonics of the Mona Passage. Transactions of the of Geophysical Research, 64, 1545-1555. American Geophysical Union, 56, 451. 78Todd, R. & Low, D. 1979. Smaller foraminifera from deep 84Weiland, T.J. 1988. The petrology of the volcanic wells on Puerto Rico and St. Croix. U.S. Geological rocks in the Lower Cretaceous formations of Survey, Professional Paper, 863, 32 pp. northeastern Puerto Rico. Unpublished Ph.D. 79Van Fossen, M.C., Channell, J.E.T. & Schellekens, J.H. thesis, University of North Carolina, 265 pp. 1989. Paleomagnetic evidence for Tertiary antic lock- 85Whetten, J.T. 1966. Geology of St. Croix, U.S. Virgin wise rotation in southwest Puerto Rico. Geophysical Islands: in Hess, H. H. (ed), Caribbean Geological Research Letters, 16, 819-822. Investigations. Geological Society of America Memoir, 80Volckmann, R.P. 1983. Upper Cretaceous stratigraphy of 98, 177-239. southwest Puerto Rico: a revision. U.S. Geological

165

CHAPTER 9

The Lesser Antilles

GEOFFREY WADGE

N.E.R.C. Unit for Thematic Information Systems, Department of Geography, University of Reading, P.O. Box 227, Reading, RG6 2AB, England

INTRODUCTION The positive anomalies, which are present after both free-air and Bouguer corrections, are mainly due to the relatively THE LESSER Antilles are a chain of islands that extend in an high crustal level of dense igneous rocks. The negative 850 km arc from Sombrero in the north to Grenada in the anomalies, in contrast, are due to the relatively great thickness south (Fig. 9.1). They are separated from the Greater An- of low density sedimentary rocks in the basins of the forearc tilles by the Anegada Passage whilst the Venezuela conti- and the accretionary wedge. nental margin starts just south of Grenada. The Lesser Antilles is a true island arc; islands built largely by volcan- Seismicity 16 ism above a subduction zone . This involves the subduc- Two types of earthquake are recognised in the arc: tion of Atlantic ocean floor westwards beneath the tectonic earthquakes associated with the subduction of the Caribbean Plate at the rate of about 20 mm/year. Compared to oceanic crust and its effects in the forearc; and volcanic island arcs globally the Lesser Antilles arc is relatively earthquakes associated with the rise of magma into the crust short, long-lived (at least since the Eocene), of only moderate beneath the volcanoes. It is possible to map the crude shape of vigour and displays a very well-developed accretionary the subducted slab from the hypocentres of the deeper wedge in its southern part. tectonic earthquakes received by seismometers from the network within the Lesser Antilles19 (Fig. 9.1 c). This shape, or Wadati-Benioff zone, shows three segments: STRUCTURE AND GEOPHYSICS (A) A vigorous northern segment (from Martinique north There are four important components to the structure of the wards) dipping about 50-60° to the WSW and reaching arc: the Atlantic oceanic crust; the forearc (upon which the depths up to 210 km. accretionary wedge is built); the arc massif itself; and the (B) A weakly-seismic southern segment (as far south as the Grenada Basin behind the arc. There is considerable along- Grenadines) dipping about 45-50° to the WNW and arc variation in this pattern. In particular, a well-defined reaching depths of up to 170 km. trench, the Puerto Rico Trench, exists only in the north. This is (C) An essentially vertical segment south of (B) extending about 250 km from the volcanic front (position of the active to the Paria Peninsula of Venezuela. volcanoes), but the distance to the equivalent feature in the south, the deformation front of disturbed accreted From the energy released by the major historical earth- sediments, is about 420 km. Further, the back-arc Grenada quakes the equivalent rates of slip and hence convergence of Basin extends only as far northwards as Guadeloupe. the subduction process can be calculated15. These rates (5 mm/year and 1 mm/year) for the northern and southern Gravity Anomalies segments are significantly less than the 20 mm/year pre- There is an arcuate pair of gravity anomalies associated dicted from global plate tectonic models. This implies that with the Lesser Antilles: positive over the arc massif and 20 much of the slip between the Caribbean and American plates is negative over the forearc or edge of the Caribbean plate . achieved without friction, that is, aseismically. This is

167 The Lesser Antilles

Figure 9.1. The Lesser Antilles region, (a) Bathymetric map. l=Volcanic Caribbees; 2=Limestone Caribbees; 3=axis of the inner arc; 4=axis of the outer arc; 5=deformation front. Isobaths in m. (b) Schematic crustal structure of the Lesser Antilles arc. l=outer forearc crust; 2=inner forearc crust; 3=arc massif; 4=oceanic basement ridges; 5=late Jurassic -early Cretaceous Atlantic crust. AA', BB', CC’, DD'=lines of sections shown in Figure 9.2. (c) Isobaths (in km) of the Benioff zone beneath the arc (after Wadge and Shepherd19). The diagonal shading shows the vertical part of the southern zone beneath the Venezuelan continental shelf. (After Maury et al.9).

particularly so in the south, where some sediment can be lacking. One feature that is certain is that to the north of St. seen on seismic reflection profiles to be subducted beneath Lucia the arc massif bifurcates into two: an Eocene to the accretionary wedge. Miocene eastern arc; and a Pliocene to Recent western arc. The reason for the distinct break in the Wadati-Benioff zone beneath Martinique and St. Lucia is not known. It Seismic Refraction and Reflection Data may represent the boundary between the North American The crustal structure of the arc down to the MOHO and South American plates, but evidence supporting this is (about 30 km depth below Martinique) has been determined

168 GEOFFREY WADGE by seismic refraction experiments3. Seismic reflection data with an isotopic age of 145-150 Ma, and younger metavol- has allowed the structural and sedimentary history of the canic rocks, some of which have interbedded cherts with a sedimentary basins within the arc to be elucidated14. There biostratigraphic age of about 112-119 Ma. Many of the are three main crustal layers: submarine escarpments of the northeast-facing part of the arc from Guadeloupe to Anguilla have revealed Upper Cretaceous (A) An upper layer of average seismic velocity of 3.3 km/s rocks in dredge hauls. Similarly, Upper Cretaceous rocks, comprising a mixture of sediments and volcanics. some with island arc affinities, are known from dredge (B) A middle layer of average seismic velocity of 6.2 km/s hauls along the Aves Swell that runs N-S behind the Lesser indicative of intermediate plutonic rocks. Antilles. The Aves Sw ell is interpreted to be an island arc that (C) A lower layer whose average velocity of 6.9 km/s was active prior to the Lesser Antilles arc. suggests basic rocks, probably oceanic crust or basic plutonic cumulates. Barbados Barbados has a unique character in the arc (see Speed, There is considerable lateral variation in the depth of this volume). It is the exposed top of the accretionary wedge these layers which is probably the result of a long and of sediments that have been scraped off by subduction, the complex history of subduction zone processes. Cross-sec- Barbados Ridge. Most of the island is covered by terraces tional models of the structure of the arc based on gravity of Pleistocene reef limestones that record rapid, differential anomalies constrained by seismic refraction and reflection uplift along an axis trending NE-SW. However, on the data show distinctive differences between the northern and northeast coast two groups of older accretionary rocks are southern segments (Fig. 9.2). In the south the volcanic front exposed. These are: (i) the Miocene-Middle Eocene Oce- sits on a single arc massif, whereas in the north the present- anic Group comprising mainly lightly deformed abyssal day volcanoes are on the western flank of the massif and rocks; and (ii) the Lower-Middle Eocene Scotland Group have contributed relatively little to its growth. In the north comprising deformed terrigenous and hemipelagic rocks models suggest a complicated forearc structure, perhaps (radiolarites). The source of the terrigenous Eocene rocks involving an additional slice of oceanic crust between the was debris flows of material containing metamorphic min- Caribbean forearc and the subducting slab. In the southern erals from South America either in a deep-sea trench or a Grenadines the high velocity layer is very shallow, which submarine fan. has led to speculation that the oceanic crust of the Grenada Basin may extend across to the forearc at this latitude. The Limestone Caribbees The islands of the northeastern Lesser Antilles from Anguilla to Marie Galante are characterised by Cenozoic STRATIGRAPHY AND VOLCANISM limestones, hence their name of Limestone Caribbees8. The limestones are of different ages; Eocene on St. Barthelemy, There is evidence of an island arc in the Lesser Antilles for Oligocene on Antigua, Miocene on Anguilla and St. Martin, 4,9 at least 50 Ma (Fig. 9.3). However, much of this evidence is and Pliocene-Quaternary on Barbuda, Grande Terre fragmentary, sometimes consisting of small regions of (Guadeloupe) and Marie Galante. Underlying volcanic outcrop whose significance to the whole arc's history is rocks representing the older, eastern branch of the arc are open to speculation. The value of stratigraphic correlation well-exposed on St. Martin, St. Barthelemy (Eocene and 7 is weak when most of the rocks are of volcanic origin and Oligocene age) and Antigua (Oligocene age). On Anguilla have been generated from essentially point sources (volca- Paleocene/Eocene age sediments overlie volc anic rocks of noes). unknown age. Volcanic rocks are not exposed in Barbuda, Grande Terre (Guadeloupe) or Marie Galante. Barbuda is Basement Exposures offset from the axis of the old volcanic front and may not Although there is no evidence for an east-facing island have been a volcanic centre. In general, the age of the oldest arc prior to the middle Eocene, it seems clear that the exposed rocks in the Limestone Caribbees increases north- basement on which the Cenozoic arc is built is not simply wards. The lack of volcanic intercalations in the Upper Caribbean oceanic oust alone. Cretaceous/Paleocene is- Oligocene Antigua Formation suggests that volcanism had land arc rocks may underlie much of the arc from Guade- ceased by this time in the Limestone Caribbees. It resumed loupe northwards. The most important of the basement millions of years later much further west. exposures is on La Desirade, east of Guadeloupe1. Beneath a plateau of Lower Pliocene limestones is a Mesozoic igne- The Volcanic Caribbees ous complex consisting of trondhjemite/rhyolite lava flows, The present day volcanic front extends from Grenada

169 The Lesser Antilles

Figure 9.2. Cross sections of the crust of the arc derived from gravity modelling constrained by seismic refraction and reflection data. The lines of section are shown in Figure 9.1b. The position of the active volcanic arc is shown by triangles. In section AA' it lies at the edge of the crustal thickening associated with the main body of the arc. The inactive volcanic arc in section DD' is contiguous with the active arc and joins it south of Grenada. Numbers give the densities (10-3 x kg.m-3) of the layers used in the gravity models: 1.8-2.55, sediment; 2.3-3.05, igneous crustal rocks; 3.3, mantle. Position of MOHO shown by dotted lines in sections AA', BB', from models including gravitational attraction of subducted lithosphere. (After Maury et al.9).

170 GEOFFREY WADGE

Figure 9.3. Map of the Lesser Antilles island arc showing the ages of the exposed rocks and the positions of the volcanic front during the Pleistocene (solid line), Pliocene (dotted line, for the central segment) and Eocene-Oligocene (crosses). The dashed line is the 200 m isobath. Inset is a bar chart showing the approximate duration of the known volcanism at each island together with the number of deep-sea ash layers from the Miocene-Recent The numbers refer to; l=Mount Noroit, 2=St. Martin, 3=St. Barthelemy, 4=Luymes Bank, 5=Saba, 6=St Eustatius, 7=St. Kitts, 8=Nevis, 9=Antigua, 10=Redonda, 1 l=Montserrat, 12=Basse Terre (Guadeloupe), 13=Les Saintes, 14=Dominica, 15=Martinique, 16=St. Lucia, 17=St Vincent, 18=Bequia, 19=Mustique, 20=Petit Canouan, 21=Canouan, 22=Mayreau, 23=Union, 24=Carriacou, 25=Ronde-Kick 'em Jenny, 26=Grenada (after Wadge 17).

to Saba and represents the locus where magma rising from back to the Eocene, but to the north the oldest exposed rocks melting induced in the asthenospheric wedge above the of the Volcanic Caribbees are of early Pliocene or latest subducting American plates reaches the surface. It appears Miocene age. that in the Pliocene submarine volcanism extended the front The southernmost part of the arc, Grenada and the about 110 km northwest of Saba to the Noroit seamount2. Grenadines archipelago, is distinctive in two ways. Firstly, From Martinique southwards the history of volcanism dates it consists of a NNE-SSW trending shallow bank, the Grena-

171 The Lesser Antilles

Figure 9.4. Isopach map of the Roseau airfall tephra in cores from the western Atlantic and eastern Caribbean. Solid (?) circles indicate cores which contain Roseau airfall ash. Upper number corresponds to core number, lower value is thickness of airfall tephra layer in on. Open (o) circles indicate cores which do not contain Roseau tephra. Shaded area represents the distribution of the subaqueous pyroclastic debris flow deposit in the Grenada Basin (after Carey and Sigurdsson6).

dines Bank, on which rocks representing much of the Ceno- source of magma with time17. Large, mature volcanoes such zoic are exposed. Secondly, alkaline Mg-rich basaltic rocks as Soufriere (St. Vincent) and Morne Diablotins (Dominica) were common in the Pliocene-Quaternary. These form the consist of a central core of domes, lava flows, ash flow and only historically active volcano, the submarine Kick'em fall deposits, with a surrounding apron of reworked sedi- Jenny north of Grenada. The oldest rocks in this part of the mentary units mainly derived from a variety of primary arc are Middle Eocene, possibly older, basalts, which may deposits. These reworked deposits (the geomorphological represent Eocene oceanic crust continuous with that in the term is 'glacis') are often very similar to each other and of Grenada Basin. Sedimentary rocks are abundant in this small lateral extent. This makes it very difficult to map any area, ranging from the Middle Eocene to Middle Miocene meaningful stratigraphy on the lower flanks of such volca- in Grenada, Carriacou and Canouan. They are charac - noes. teristically turbiditic or pelagic in character and appear to Much more useful for stratigraphic purposes are the represent gravity flow deposits in small basins that shoaled primary pyroclastic deposits. Prior to the start of wide- with time. During the late Oligocene to early Miocene these spread European colonisation of the Lesser Antilles around sedimentary rocks were variably deformed by faulting, fold- 1650 A.D., the islands had been occupied intermittently for ing and uplift in a compressional regime. about 3000 years by Amerindian peoples. Their shell mid- From St. Vincent to Basse Terre (Guadeloupe) the arc dens and pottery remains can provide a useful marker for consists of large islands created by overlapping volcanic dating the upper age limit of overlying pyroclastic deposits, massifs. The rocks of St. Vincent and Basse Terre are as in St. Kitts and Martinique. Hot pyroclastic deposits can younger than 5 Ma, though there may be older, buried rocks. carbonize wood and radiocarbon dating techniques have The age of the rocks decreases systematically along the axis been used extensively to decipher pyroclastic volcanic stra- of these two islands, northwar ds in St. Vincent and south- tigraphy back to about 40,000 yr B.P., particularly thai of wards in Basse Terre, suggesting migration of the local Mount Pelee on Martinique. Many ash fall deposits and

172 GEOFFREY WADGE

Figure 9.5. Cross-sections along the Lesser Antilles volcanic front The upper section shows the crustal structure from Grenada to Guadeloupe determined from the seismic refraction results of Boynton et al.3. The base of the crust is shown (dash-dot line) beneath Martinique. The large dashed line is the top of the lower crustal layer (average seismic velocity=6.9 km/s). The solid line is the more accurately defined top of the upper crustal layer (average seismic velocity=6.2 km/s). The small dashed line is this same boundary derived from the time-term solutions of separate line segments rather than for the whole survey (solid line). Vertical exaggeration 10:1. The lower section, at true scale, illustrates the segmentations of the volcanic front and the concept of plume traces. The dotted line is the mean position of the Benioff zone from Wadge and Shepherd19, and the two bars of horizontal ruling shows the positions of the inferred breaks in the subducting slab. The bars of stippling represent the plume traces in section and are labelled alphabetically; GhGrenada, J=Ronde/Kick 'em Jenny, C=Carriacou, PC=Petit Canouan, B=Bequia, V=St Vincent, L=St Lucia, M=Martinique, D=Dominica, LS=Les Saintes, Gu=Basse Terre (Guadeloupe), Mo=Montserrat, R=Redonda, NK=Nevis/St. Kitts, E=St. Eustatius, S=Saba, L=Luymes Bank. Volcanoes active during the last 0.1 Ma are shown as black lines and are numbered; 1-Mt St. Catherine, 2=Kick 'em Jenny, 3=Soufriere, St. Vincent, 4=Qualibou, 5=Mount Pelde, 6=Mome Patates, 7= Micotrin/Morne Trois Pitons, 8=Morne Diablotins, 9=Soufriere, Guadeloupe, 10=Soufriere Hills, 1l=Nevis Peak, 12=Mt. Misery, 13=The Quill, 14=The Mountain. The horizontal arrows show the sense of migration of the three proposed examples of lateral motion of plume traces. The two inset histograms are of the frequency of the spacings between the centres of plume traces and of volcanoes (<0.1 Ma). (After Wadge17).

173 The Lesser Antilles

large pyroclastic flows reach the sea, but only the largest centres such as the Qualibou volcano in St. Lucia Rhyolites reach the deep basins such as the Grenada Basin. By pis- are quite rare and are mainly differentiates of tholeiit ic ton-coring the sediments in these basins at numerous sites magmas. around the arc these distal deposits can be identified13. No major plutons are exposed. Small, high-level intru- Using biostratigraphic dating together with trace element sions up to 1 km in diameter are known from, for example, analysis of the ash layers, the age, source, extent and volume St. Barthelemy (Eocene) and Mustique (Oligocene), but the of the ash deposits can sometimes be determined. It was in dioritic plutons that are presumed to underlie the main this way that the source of the largest eruption in the arc islands are still deeply buried. Plutonic nodules are carried during the last 100,000 years was determined6. Submarine to the surface during eruptions, but to gain an insight into debris flow deposits to the west in the Grenada Basin and what must lie 5-10 km below volcanoes such as Soufriere, widespread air fall ash to the east were correlated geochemi- St. Vincent, we must look to older, eroded island arcs such cally and temporally. These deposits represent different as the mid-Cretaceous diorite-gabbro pluton of Tobago (see facies of the same eruptive deposits and can be most closely Jackson and Donovan, this volume). correlated with ignimbrite deposits on Dominica from the The geochemistry of the younger volcanic rocks allows a Micotrin volcano that are dated about 30,000 yr B.P. (Fig. four-fold subdivision5,9: 9.4). Eastern Martinique and St. Lucia contain important (A) Alkalic/subalkalic basalts. Mainly found in Grenada, the exposures of Miocene rocks, which are the most northerly Grenadines and southern St. Vincent, these rocks are in the Volcanic Caribbees. Upper Miocene (11-5 Ma) rocks rich in Mg, Cr and Ni. of both calc-alkaline and arc-tholeiite character occur along (B) K-poor series (0.5% < K2O at SiO2 = 50%). Typically the length of St. Lucia with the oldest (18-15 Ma) examples represented by island arc tholeiites in St. Kitts, St. being found in the north. Oligocene age rocks are known Eustatius and St. Vincent. from the St. Anne and Caravelle peninsulas of Martinique, (C) Medium-K series (0.5% < K2O < 0.9% at SiO2 = 50%). but the bulk of eastern Martinique displays a remarkable The most voluminous products of the arc, typically record of almost continuous migration of volcanic activity andesite, common in Montserrat, Grande-Terre from tholeiitic products in the east at about 16 Ma to calc - (Guadeloupe), Dominica and Mount Pelee, Martinique. alkaline rocks in the west at about 6 Ma. This is the only (D) High-K series (K2O > 0.9% at SiO2 = 50%). Known place in the arc where direct evidence for the transition from only in the central and southern parts of the arc (for the eastern Limestone Caribbees front to the current example, Pitons du Carbet (Martinique), Grenada). Volcanic Caribbees front has been observed. The northern segment of the arc from Montserrat to Overall there is a definite pattern of increasing amounts of Saba consists of smaller islands whose volcanoes reach to incompatible elements and radiogenic strontium in the lower altitudes than the volcanoes of the central part of the volcanic rocks from north to south. This has been correlated arc. Volcanic output is also much less. One effect of the with an increasing contribution to the asthenospheric melt lower altitude of these volcanoes is more intimate interac- by subducted sediments from the Orinoco. Superimposed tion with the sea. For instance, at St. Eustatius the last on this is the tendency at any one volcano for products to volcanic activity has built up a cone, the Quill, off the evolve from K-poor to K-rich with age. southern coast of the earlier volcanoes. On the southern side of the cone a steeply-dipping bed of Pleistocene limestone, tuff and gypsum, the White Wall Formation, is exposed. This is ACTIVE VOLCANOES thought to be a part of the shallow, submarine shelf bodily forced up from beneath the sea by a rising mass of partly There are no hard and fast rules about whether a volcano is solidified magma, or cryptodome, within the body of the active, that is, still capable of erupting, or not. Figure 9.5 shows 14 central volcanoes that have erupted repeatedly in main cone. 10 the last 100,000 years in the Lesser Antilles , though fiiture work may revise this figure. Dominica has at least three PETROCHEMISTRY active volcanoes and is the most active island in the arc, though the least well-known volcanologically. However, The commonest rock type on the islands of the Lesser compared to many other island arcs the Lesser Antillean Antilles is probably a reworked conglomerate of andesitic volcanism is not particularly vigorous. There are only four composition. Basalts are well-represented throughout the volcanoes that have erupted new magma during the recorded arc. Dacites are more localised, but important in major history of the last 300 years: Kick 'em Jenny; Soufriere (St.

174 GEOFFREY WADGE

Figure 9.6. Maps showing the morphological limits of Mount Pelee (Martinique) and Soufriere (St. Vincent), with the distributions of the products of their eruptions during this century. (After Roobol and Smith11).

Vincent); Mount Pelee (Martinique); and Soufriere at the volcanoes of the arc during the historic period; phrea- (Guadeloupe). The near -simultaneous eruptions of Soufri- tic eruptions and seismic crises. Phreatic eruptions are ere (St. Vincent) and Mount Pelee in 1902 caused a sensa- usually small explosive eruptions involving steam and old tion when about 30,000 people were killed by pyroclastic rock fragments. The eruption of Soufriere (Guadeloupe) in surges, most in the town of St. Pierre on the southwestern 1976-1977 which caused the evacuation of the town of flank of Mount Pelee. Subsequent study of the eruptive Basse Terre for several weeks was of this type, probably behaviour and deposits of both volcanoes has made them caused by rising magma within the volcano heating the type examples of basaltic pyroclastic flow and dome/crater groundwater. Seismic crises are even more common and lake growth (Soufriere), and andesitic block-and-ash flows typically involve tens to hundreds of small earthquakes at (Mount Pelee)11 (Fig. 9.6). depths of a few kilometres as rising magma deforms the There are two other types of activity that have occurred basement of the volcanoes.

175 The Lesser Antilles

HAZARDS FROM FUTURE ERUPTIONS material needed by government to act in time of crisis. It is a problem common to many volcanoes, but particularly to The most regular natural hazard faced by the people of the those of the Lesser Antilles which erupt infrequently, that Lesser Antilles are hurricanes. Meteorological tracking and hazard assessments and disaster preparedness plans can be telecommunications nowadays provide considerable lead ignored unless the hazard is perceived to be imminent11. time for people to take cover and secure property. The With hindsight, the town of St. Pierre on the flanks of situation for the hazards from earthquakes and volcanic Mount Pelee was (and is) seen to be in a location at great eruptions is rather different. Earthquakes rarely provide potential hazard from pyroclastic flows. A similar conclusion precursory symptoms and the only practical steps to mitigate can be reached for some other major towns in the Lesser Antilles their effects is to design buildings to withstand shaking. such as Plymouth (Montserrat) and Basse Terre However, erupting volcanoes usually do provide plenty of (Guadeloupe). precursory symptoms prior to eruption. The difficulty here lies in interpreting them in terms of the probable course of ACKNOWLEDGEMENTS—I thank the Geological Society of America and events. R.C. Maury for permission to reproduce Figures 9.1 and 9.2; S.N. Carey and Elsevier for permission to reproduce Figure 9.4; and M.J. Roobol and Springer There are three separate, but dependent, tasks facing Verlag for permission to reproduce Figure 9.6. scientists and government officials who must evaluate hazards from future eruptions: (1) monitoring the state of activity; (2) assessing the probability of hazards around the REFERENCES volcano; and (3) devising a plan of action based on that assessment. Individual island governments generally do not 1Bouysse, P., Schmidt-Effing, R. & Westercamp, D. 1983. La have the expertise to carry out the first of these tasks. Desirade islands (Lesser Antilles) revisited: Lower Regionally-based organisations such as the Seismic Re- Cretaceous radiolarian cherts and arguments against an search Unit of the University of the West Indies, based in ophiolitic origin for the basal complex. Geology, 11, Trinidad, can fill this role, primarily by monitoring the 244-247. seismicity beneath the volcanic centres. Ideally there 2Bouysse, P. & Westercamp, D. 1990. Subduction of Atlantic should be a baseline set of measurements to establish what aseismic ridges and late Cenozoic evolution of the is 'normal' for each volcano. Such measurements may in- Lesser Antilles island arc. Tectonophysics, 175, 349-380. clude seismicity, geodetic shape of the volcano, and gas 3Boynton, C.H., Westbrook, O.K., Bott, M.H.P. & Long, temperature and composition. These are best carried out R.E. 1979. A seismic refraction investigation of the from local observatories, which ar e unfortunately expensive crustal structure beneath the Lesser Antilles island arc. to maintain. Only two have operated in recent years, on Royal Astronomical Society Geophysical Joumal, 58, Soufriere (Guadeloupe) and Soufriere (St. Vincent). 371-393. When a volcanic crisis begins there is a need to greatly 4Briden, J.C., Rex, D.C., Faller, A.M. & Tomblin, J.F. 1979. K- increase the level of observational measurement. This ne- Ar geochronology and palaeomagnetism of volcanic rocks cessitates bringing in equipment and expertise. The inter- in the Lesser Antilles island arc. Philosophical national volcanological community has shown itself Transactions of the Royal Society of London, A291, capable of responding to such needs as in St. Vincent in 485-528. 1979. After many months of mild precursory activity at 5Brown, G.M., Holland, J.G., Sigurdsson, J., Tomblin, J.F. & Soufiiere (St. Vincent), a rapidly-assembled team of scien- Arculus, R.J. 1977. Geochemistry of the Lesser tists from Trinidad, the USA and UK deployed a monitoring Antilles volcanic island arc. Geochimica et Cosmo- network and advised on the safe evacution of 15,000 people chimica Acta, 41,785-801. prior to the most dangerous explosive phase of the erup- 6 12 Carey, S.N. & Sigurdsson, H. 1980. The Roseau ash: deep- tion . sea tephra deposits from a major eruption on Dominica, The observational data obtained before and during a Lesser Antilles arc. JournalofVolcanology and developing crisis (which may last many weeks) must be Geothermal Research, 7, 67-86. interpreted in terms of the location, time and severity of 7Lewis, J. and Robinson, E. 1976. A revised stratigraphy and imminent eruptions. The interpretation must also be put into geological history of the Lesser Antilles: in Causse, R. the context of the hazardous consequences of these erup- (ed.), Transactions of the Seventh Caribbean Geological tions. These hazards are best as sessed before the crisis 18 Conference, Guadeloupe, 30th June-12th July, 1974, 339- occurs by study of the deposits of earlier eruptions . Sci- 344. entific advice on the eruptive status of a volcano, together 8Martin-Kaye, P.H.A. 1969. A summary of the geology of with maps of zones at risk from potential hazards are the

176 GEOFFREY WADGE

the Lesser Antilles. Overseas Geology and Mineral national, Woods Hole. Resources, 10, 172-206. 15Stein, S., Engeln, J.F. & Wiens, D.A. 1982. Subduction 9Maury, R.C., Westbrook, O.K., Baker, P.E., Bouysse, P. & seismicity and tectonics in the Lesser Antilles. Journal Westercamp, D. 1990. Geology of the Lesser Antilles: of Geophysical Research, 87, 8642-8664. in Dengo, G. & Case, J.E. (eds), The Geology of North 16Tomblin, J.F. 1975. The Lesser Antilles and Aves Ridge: America. Volume H. The Caribbean Region, 141-166. in Nairn, A.E.M. and Stehli, F.G. (eds), The Ocean Geological Society of America, Boulder. Basins and Margins. Volume 3. The Gulf of Mexico and 10Robson, G.R. & Tomblin, J.F. 1966. Catalogue of the the Caribbean, 467-500. Plenum, New York. Active Volcanoes of the World: Part 20, West Indies. 17Wadge, G. 1986. The dykes and structural setting of the International Association for Volcanology, 56pp. volcanic front in the Lesser Antilles island arc. Bulletin 11Roobol, M.J. & Smith, A.L. 1989. Volcanic and associ- Volcanologique, 48, 349-372. ated hazards in the Lesser Antilles: in Latter, J.H. (ed.), 18Wadge, G. & Isaacs, M.C. 1988. Mapping the volcanic Volcanic Hazards, 57-85. Springer-Verlag, Berlin. hazards from Soufriere Hills volcano, Montserrat, 12 Shepherd, J.B., Aspinall, W.P., Rowley, K.C., Pereira, West Indies using an image processor. Journal of the J.A., Sigurdsson, H., Fiske, R.S. & Tomblin, J.F. 1979. Geological Society, London, 145, 541-551. The eruption of Soufriere volcano, St. Vincent, April- 19Wadge, G. & Shepherd, J.B. 1984. Segmentation of the June 1979. Nature, 282, 24-28. Lesser Antilles Subduction zone. Earth and Planetary 13 Sigurdsson, H., Sparks, R.S.J., Carey, S.N. & Huang, T.C. Science Letters, 71, 297-304. 1980. Volcanogenic sedimentation in the Lesser An- 20Westbrook, O.K. 1975. The structure of the crust and tilles arc. Journal of Geology, 88, 523-540. upper mantle in the region of Barbados and the Lesser 14 Speed, R.C. & Westbrook, O.K. 1984. Lesser Antilles arc Antilles. Royal Astronomical Society Geophysical and adjacent terranes. A tlas 10, Ocean Margin Drlling Journal, 43, 201-242. Program, Regional Atlas Series. Marine Science Inter-

177

CHAPTER 10

Barbados and the Lesser Antilles Forearc

ROBERT C. SPEED

Department of Geological Science, Northwestern University, Evanston, Illinois 60208, U.SA.

INTRODUCTION SOUTHERN LESSER ANTILLES FOREARC

THE ISLAND of Barbados differs geologically from other The southeastern Caribbean region is comprised of two, islands of the eastern Caribbean by virtue of its evolution in possibly three, major plates, the Caribbean, South American an accretionary prism rather than a magmatic arc. The and, possibly, North America. The position of the South accretionary prism is the extensive Barbados Ridge (Fig. American-North American plate boundary within the Atlan- 10.1), which is east of and separate from the ridge of the tic basin is uncertain. Because the relative motion between Lesser Antilles magmatic arc platform. Barbados island is the North and South American Plates is small near Bar- the only exposure above sea level of the Barbados Ridge. bados1, they can be considered a single plate, the American The Barbados Ridge lies above the subduction zone Plate (Fig. 10.2). The Caribbean-American plate boundary between the descending Atlantic oceanic lithosphere of the is well-defined from St. Lucia north by earthquakes that American Plate and the overriding Caribbean Plate (Fig. define a westward-dipping subduction zone. South from St. 10.1). The accretionary prism has evolved over 50 Ma by Lucia, the plate boundary probably continues south below the processes of accretion, deformation and diapirism. Bar- Barbados (Fig. 10.2), according to the locus of minimum bados island permits the study of the rocks, ages of events Bouger gravity anomalies, but south of there its position is and tectonic structures of the inner realm of the prism, and increasingly uncertain. is therefore a major source of information on the geologic The Caribbean-American plate boundary of the south- evolution of the Barbados Ridge as a whole. Because the eastern Caribbean is overlain by the Lesser Antilles arc Barbados Ridge is technically oceanward or forward of the system, which comprises a magmatic arc platform and an Lesser Antilles magmatic arc, it is included in the Lesser extensive forearc (Fig. 10.1). The forearc consists of two Antilles forearc, together with the forearc basins, such as the principal elements, the Barbados accretionary prism and Tobago Trough and the Lesser Antilles Trough (Fig. 10.1). forearc basins, of which the Tobago Trough is situated at the An atlas of geophysical and geological data from the Lesser latitude of Barbados. The accretionary prism lies above Antilles magmatic arc and forearc has been published by incoming Atlantic oceanic lithosphere of the American Plate Speed et al28. between an outer deformation front (ODF; Fig. 10.1), the This chapter discusses first the structure of the Lesser locus of current accretion of sediment, and a transition zone Antilles forearc as a whole and then presents the geology of to undeformed forearc basin strata. The transition zone, also Barbados island as a sample of that of the forearc. It then the site of active deformation, is called the Inner Forearc returns to the tectonic evolution of the forearc in the light of Deformation Belt (IFDB; Fig. 10.1). Barbados island is at Barbadian geology. Three sequential events in the history the structural high, the thickest part of the prism, and prob- of the island of Barbados are given particular emphasis: ably above the trace of subduction between crystalline early accretion in the construction of the Barbados Ridge; American and Caribbean lithospheres. late stage deformation that affected the inner region of the The profile width and maximum thickness of the prism forearc; and the tectonic rise of Barbados above sea level in below Barbados are about 250 km and 20 km, respectively the last million years. (Fig. 10.1B). The prism tapers eastward from the structural high because of progressively lesser duration for thickening

179 Barbados and the Antillean Forearc

Figure 10.1. Regional tectonic setting of Barbados. (A) Southern Lesser Antilles arc system. Key: T.P =Tobago platform; T=Trinidad; F.T.B.=fold-thrust belt; S.A.=South America. (B) Map showing forearc divisions; Vs are active volcanoes. (C) Cross-section of forearc along trace AA' on (B). by the combined mechanisms of frontal accretion, under - mation is taken up by out-of-sequence thrusting on shal- plating, backrotation, horizontal contraction and sedimenta- lowly west-dipping faults36 (Fig. 10.1C). The lower section tion at the sea bed. The prism taper west of the structural of incoming sediment underrides the ODF and the accretion- high mainly reflects the tectonic wedging of the prism west ary prism (Figs 10.1C, 10.3). The distance of underriding into the forearc basin31 (Fig. 10.IB). The profile width and before attachment to the base of the accretionary prism volume of the prism increase from north to south with (underplating; Fig. 10.1C) is the full prism width north of 36 proximity to South American sources of terrigenous sedi- 15.5°N, according to deep seismic sounding , but is uncer - ment. North of 15°N, the prism is probably mud-rich14,15, tain at the latitude of Barbados. The concept that thick whereas it is probably increasingly sand-rich to the south35 . underplating has occurred below Barbados (Fig. 10.1 C) has The ODF is the locus of the most seaward emergent thrust been proposed as one possible explanation for extensive 23 in an imbricate stack above the prism's basal detach- late-stage deformation of Barbados and the IFDB . ment fault, which is less than halfway down through the The forearc basin of the Tobago Trough west of Bar- section of incoming sediment on the Atlantic litho- bados (Fig. 10. IB) contains as much as 12 km of unde- 14,15,37 formfed sediment above oceanic crust that is probably of sphere (Fig. 10.3). The transect of DSDP leg 78A-ODP 26,27 leg 110 drillholes (Fig. 10.1) across the ODF at 15.5°N middle Eocene age (about 50 Ma) . The sedimentary (Fig. 10.IB) indicates that the thrust packets shorten hori- succession thickens southeastward as a wedge off the down- tilted eastern flank of the arc platform to maxima along the zontally and thicken by internal folding and backrotation 26 westward from the ODF. A few tens of kilometres west- Inner Deformation Front (IDF; Fig. 10. IB). Deformed ward, such contractile mechanisms lode and further defor- forearc basin strata can be tracked east of the IDF by seismic

180 ROBERT C. SPEED

Figure 10.2. Map of southeastern Caribbean showing possible geography of Caribbean and American plates. CST is subduction trace between crystalline lithospheres of the two plates below terranes and accreted sediments whose surface extent is delimited by the outer deformation front (ODF) and inner deformation front (IDF); hachures indicate locus of Neogene island arc volcanoes. reflection sections (Fig. 10.4), but, except for shallow inter- mately the last 20 Ma. The subprism detachment probably vals, isopachs are lost at the IDF31. Within the Tobago developed above crystalline basement close to and above Trough, strata are interpreted to consist of calcareous pe- the subduction trace (Fig. 10.1C). A later section returns to lagic, and South American hemipelagic and quartzose, tur- the evolution of the IFDB in the light of the geology of biditic sediments in the Paleogene section. The Neogene Barbados. section is interpreted to contain upwardly increasing amounts of volcanogenic sediment and ashfall from the Lesser Antilles magmatic arc platform, whose emergence TECTONOSTRAT1GRAPHY OF BARBADOS and volcanism began at about 20 Ma. The IFDB is a belt of active deformation about 50 to 70 Barbados island (Fig. 10.5) provides by outcrop and drilling a study volume of the structural high (about 1000 square km wide to the east of the IDF (Fig. 10.1B). The belt extends 2,3,7-10,16,17,21-23,25,33,34 at least 400 km on strike between about 12°N and an km) . These studies indicate that some undefined northern limit. The IFDB is a fold and thrust belt mappable rock units at the surface and in the subsur- developed in thick strata of the Tobago Trough by arcward face are in tectonic contact, and probably never evolved as (westward) wedging of the accretionary prism in approxi- an initial depositional succession. Rather, the rock units

181 Barbados and the Antillean Forearc

Figure 10.3. Detail of structure at modern accretionary front of Barbados accretionary prism at DSDP-ODP sites of Figure 10. IB, from Moore et al (1988); bathymetry in m; seismic section (CRV 128) has vertical exaggeration of 3; trace shown in black line on bathymetric map; interpreted depth section shown at bottom without vertical exaggeration. have been brought together at various times from relatively (4) Diapirs, which consist of organic mud-matrix melange. distant sites of deposition by faulting. Thus, we must con- These have been emplaced through the basal complex, sider the succession of rock units of Barbados as a tectonos - prism cover and mainly along the base of the Oceanic tratigraphy, and enquire about the times, conditions and allochthon, but also locally within it. directions of tectonic emplacement, together with the tradi- tional sequence of deposits and unconformities. The four tectonostratigraphic units are overlain by Barbados is comprised of four major tectonostrati- autochthonous reef-related deposits of Pleistocene and graphic units, which are, in order of emplacement (Figs Holocene age. The reef deposits record the progressively 10.5,10.6): broadening domal emergence of the island during glacial- eustatic oscillation of sea level6,11,13. The rise of Barbados as (1) The basal complex, which consists of deformed sedi an island, beginning before about 650 ka4, is anomalous with ments of Eocene depositional age accreted at an Eocene respect to regional submergence of the structural high in the outer deformation front, probably during the earliest late Neogene31 and is thought to be due mainly to local years of prism history. It extends to maximum well diapirism25 (see below). depths (4.5 km) and, probably, far below22,23. (2) The prism cover, which consists of strata including Basal complex mid-Miocene beds that were deposited on or are The basal complex (Figs 10.5, 10.6) is the lowest tec- parautochthonous to the basal complex2,23. tonostratigraphic unit of Barbados. It is very thick and (3) The Oceanic allochthon, which is thrust above the basal extensive, and probably forms the upper half if not the entire complex and prism cover, and consists of fault slices of thickness of the accretionary prism at its structural high (Fig. the Oceanic beds. These are Eocene to Miocene strata 10.1C). The basal complex is composed of fault-bounded as thick as 1.5 km that were thrust east from the outer packets of marine sedimentary rocks of known Eocene (eastern) flank of the Tobago Trough during Neogene depositional ages. At least 45 such packets have been iden- deformation of the IFDB25,33. tified at the surface9,22 and, almost certainly, many more

182 ROBERT C. SPEED

Figure 10.4. Line drawing of seismic reflection profile T3 across the IFDB; track shown on Figure 10.9 (from Torrini and Speed31). exist at depth. The hemipelagic suite consists of radiolarian mud- The packet boundaries are plane or smoothly-curved stones and radiolarites that are all early or early middle faults. Surface dips of such faults are between 0° and 90°. Eocene and are apparently older than the terrigenous beds. In the main, packets contain remarkably coherent internal This suite occupies discrete, thin (10-30 m) fault packets that structure, consisting of homoaxial, shallowly-plunging are apparently distributed in minor volume throughout the folds of a single generation and faults with only minor basal complex. Such packets have remarkable continuity on displacement. Packet thickness normal to the bounding strike, but their continuity with dip is uncertain. faults is between 20 and 1000 m. Some packet-bounding An admissible model of the depositional setting of faults have multiple generations of structures in their walls, rocks of the basal complex23 is that the hemipelagic suite indicating reactivation. Well inside the packet boundaries, and much of the terrigenous suite were in a stratigraphic rocks have mainly continuous folds as early structures. The succession above oceanic lithosphere fronting Eocene packets are interpreted to be products of accretion by off- northern South America (Fig. 10.7). Channels crossing the scraping (Fig. 10.3) early in the development of the Bar- abyssal plain delivered terrigenous sediments, perhaps as far bados accretionary prism. as the outer deformation front. It is not clear whether or not Rocks of the basal complex are composed of two main trench wedge deposition occurred at the ODF. The detach- lithic suites, terrigenous and hemipelagic. The terrigenous ment below the Eocene prism is thought to have propagated suite is at least 90% of the volume; it is approximately half through rocks of the hemipelagic suite because no older illitic mud, and half quartz sand and sandstone. It was strata are known in the basal complex. Further, the originally called the Scotland Series7 and Scotland Forma- hemipelagic suite probably contained the detachment be- tion20. The suite is a product of sediment-gravity flows cause its radiolarian abundance likely caused a high-poros- of varied mechanisms that brought South American ity zone in the Eocene section, as found in the modem continental sediment to the deep seafloor in the middle incoming section in the DSDP leg 78 A-110 transect15. and late Eocene10,15. Such sediments occur mainly in three different types of layer assemblies: mudstone- Prism cover predominant sequences of low porosity, with or without Prism cover comprises those sedimentary rocks that are impermeable, fine-grained sandstone, in thicknesses up to interpreted to have been deposited on the basal complex 200 m; sequences 10-100 m thick of sandy turbidites that after the complex was accreted. Therefore, the permissible are commonly matrix-rich or throroughly cemented and age range of the prism cover is late Eocene to late Neogene; have a low porosity; and medium- and coarse-grained, the latter age comes from the age of emplacement of the sandstone-rich successions as thick as 125 m in which Oceanic allochthon above the structural high. Prism cover cementation is meagre and porosity high (40-50%). Each on Barbados includes several geographically discrete type exists in many packets. There is no interpacket units2,23. Among these, the Woodbourne Intermediate Unit, stratigraphy and the total thickness of a terrigenous which occupies the Woodbourne Trough of southern Bar- succession that may have existed before accretion is bados (Figs 10.5,10.6), has by far the largest volume. Other uncertain.

183 Barbados and the Antillean Forearc

Figure 10.5. Geologic map and sections of Barbados island (Oceanic nappes are equivalent to Oceanic allochthons in text). Sites with heavy dots are deep wells. units to the north in Barbados are in fault packets within or the Woodbourne Trough (Fig. 10.5). Its stratigraphy and at the top of the basal complex, except for the Kingsley Unit facies, moreover, are heterogenous. The variability is that may be in place23. Much of the prism cover is subsur- thought to be due to deposition during late stage deformation face and its base has not been cored or closely studied. It that caused the synclinal subsidence of the trough, together cannot be proved whether such cover is autochthonous or with folding and local thrusting of the trough floor. Only not However, owing to the close sedimentologic and struc- Miocene ages have been obtained from the Woodbourne tural links with nearby rocks of the basal complex, all prism Intermediate Unit. The unit has a generalized stratigraphy, 16 cover is regarded as locally deposited (autochthonous or with lower mudstones and upper sandstones . Cores show parautochthonous). that at least some strata are turbiditic and are composed of As a whole, prism cover is of terrigenous (quartz-clay) particles mainly derived from the basal complex, including composition and commonly mud-rich. It is sporadically Eocene radiolarians and pollen, but also derived from earlier marly and locally constituted by coarse-grained, channel- prism cover. filling sandstone and debris flow. Microfossil ages in core Prism cover probably evolved as a thin, hemipelagic and outcrops of prism cover are Oligocene, Miocene and coating of the basal complex in Oligocene and possibly late possibly Pliocene2,23. Eocene time. With the onset of late stage deformation and The Woodbourne Intermediate Unit appears to have a major folding early in the Miocene (see below), the flanks highly varied thickness, from 0.2 to 1 km, along the axis of of a central anticline of Barbados (Fig. 10.8) became the

184 ROBERT C. SPEED

source of sediment gravity flows that accumulated in the by far is that indicated invading the central anticline on synclinal basins. Local surficial eruption of the melange Figures 10.5,10.6 and 10.8 (packet 18 of Larue and Speed9). diapir gave rise to at least one unit of prism cover and may The active body is continuous on strike (NE-SW) for 20 km have been a copious source of debris taken away by across the island and probably to the SW offshore an equal downslope transport. The deposition of prism cover was distance31. In its lower reaches, it is interpreted to be a ended by overthrusting of the Oceanic allochthon. vertical dyke of width at least 3 km. As noted, it is mush- roomed at higher levels below the Oceanic allochthon. Oceanic allochthon All exposures of the melange diapirs show generally The Oceanic allochthon probably fully-covered the ba- similar rocks; angular blocks of sedimentary rocks up to 10 sal complex and prism cover of Barbados upon its emplace- m long, granules of green mudstone and foliated, organic- ment in the late Miocene or Pliocene. The allochthon has rich, sandy mudstone matrix. Pollen ages for the matrix are provided an important seal to upward migration of fluids Paleocene to Middle Eocene. Block types are mainly hard and, apparently, to the upward flow of the diapiric melange. rocks of the basal complex; bitumen-cemented, coarse- The allochthon is as thick as 1.5 km in the Woodbourne grained, quartz sandstones; and light carbon nodular lime- Trough, but rarely more than 0.3 km in thickness elsewhere, stones. The source region of the melange diapirs is below due apparently to erosion before emergence and deposition maximum well depth (4.5 km). However, the thermal im- of the Pleistocene reef (Fig. 10.5). The allochthon is a series maturity of kerogens in the matrix indicates a maximum of thrust nappes in its lower 100-200 m, but is apparently a 30,33 depth of about 10 km, given a present temperature gradient single, undisrupted sheet above the nappes . The rocks of of 10° km-1 for Barbados24. The diapirism may have arisen the allochthon are the Oceanic beds (Oceanic Se- 7,18,19 by fluid overpressure at an especially impermeable thrust ries ), which are an initially continuous succession of slice or stratum within the basal complex, of which the green mainly calcareous pelagic, muddy and ashfall layers of mudstone granules may be samples. However, the bulk of middle Eocene to early Miocene age. These include modest the diapir bodies is material entrained during passage facies changes that reflect local channelization and slope through the basal complex and, perhaps, prism cover. The failure in some of the depositional sites of the Oceanic beds, instability permitting buoyant rise was caused by late-stage and that occur mainly in Upper Oligocene and Miocene folding of Barbados (Fig. 10.8). Each of the melange diapirs Oceanic beds. The depth of deposition is considered to have has probably acted as a conduit for hydrocarbon advection, been lower bathyal and abyssal from study of benthic 19,29 in addition to mass transport of organics together with foraminifers and ostracodes . The depositional site was inorganic particles. This can be inferred from the veins of a hemipelagic basin. The source of the Oceanic allochthon bitumen in the diapirs and their walls, together with the high is interpreted to have been the outer (eastern) flank of the proportion of organic carbon in the diapirs relative to that of Tobago Trough forearc basin; the tectonic transport was at the basal complex. least 20 km and probably substantially greater. The base of the Oceanic allochthon is the sub-Oceanic fault zone (SOFZ; Fig. 10.5), which is a 1 to 50 m thick zone TECTONOSTRATIGRAPHIC EVOLUTION OF BARBADOS of slices of: claystone representing carbonate-leached Oce- anic beds; slices of footwall rock; nodules of light-carbon 32 The tectonostratigraphic evolution of Barbados occurred in limestone ; and concentrated bitumens. In contrast to the two stages23. The early stage consisted of the offscraping of abundant evidence for transport of water and hydrocarbons sediments of the basal complex late in the Eocene. This and for shear strain in the SOFZ, the Oceanic beds higher in produced an assembly of fault-bounded packets that have the allochthon are almost unaffected by either phenomenon, probably been at or within 1 -2 km of the seabed ever since24. implying that the allochthon was an impermeable cap to its The late stage was constituted by the emplacement of the defluidizing substrate. Moreover, the mushrooming of the other three tectonostratigraphic units on, and of the refold- main diapir of the central anticline at the base, and the related ing and reactivated faulting of, the basal complex. On Bar- doming (Fig. 10.8), of the Oceanic allochthon, indicate that bados, the earliest dated late stage event is subsidence of the it behaved largely as a cohesive plastic membrane during Woodbourne Trough at 16 Ma23. However, if late stage postemplacement deformation. events are correctly correlated with IFDB development and Melange diapirs related changes in the Tobago Trough, then they probably began on Barbados as far back as 20 Ma (early Miocene). It Exposures below the Pleistocene reef indicate that Bar- is not clear whether or not a period of nondeformation bados contains at least five bodies of diapiric, mud-matrix occurred between the two stages of deformation. melange that are discrete at the present surface. The largest On Barbados, late stage deformation included NW-SE

185 Barbados and the Antillean Forearc

Figure 10.6. Tectonostratigraphic stacking diagram for Barbados island. Key: heavy lines indicate tectonic contacts; tight lines indicate depositional contacts; dashes indicate uncertainty; lined pattern indicates prism cover.

Figure 10.7. Map showing conceptual depositional palaeogeography of Eocene sediments now in basal complex of Barbados.

contraction, taken up by open folding and reactivation of anticline has been the locus of concentrated diapirism, faults, commonly with southeast vergence, but locally with which was either invited by or propelled the amplitude northwest backthrusting. A major syncline-anticline-syn- growth of this fold. The diapirism continues today. The cline set is thought to have developed across Barbados (Fig. northern syncline is inferred from relics of prism cover in 10.8, top) as an early and fundamental structure in such the subsurface of northern Barbados23. There, south-vergent contraction. The southern syncline, the Woodbourne thrusts cut the syncline and imbricated it complexly (Fig. Trough, has probably subsided continuously since the early 10.5). Succeeding events in the late stage deformation (Fig. Miocene, accumulating a thick, troughal unit of prism cover, 10.8) were: progressive thrusting of the basal complex and the Woodbourne Intermediate Unit (Fig. 10.8). The central prism cover; the overthrusting from the west of the Oceanic

186 ROBERT C. SPEED

Figure 10.8. Schematic serial cross-sections through Barbados showing structures produced by progressive late stage deformation; the central anticline undergoes axial diapirism; the southern syncline subsides throughout the time series. allochthon; and the rise and mushrooming below the Oce- the roof thrust, in a west-verging duplex. The eastern zone anic allochthon of the central melange diapir. The greatest contains west-thinning wedges of accretionary prism and, thickness and youngest ages of the Oceanic allochthon are between them, the easternmost slices of duplexed forearc in the Woodbourne Trough because its relative lowness and basin strata (Fig. 10.10). Syndeformational cover sediments continuing subsidence protected the allochthon from ero- occupy synclinal troughs below the seabed and probably sion. The oldest tracts of the autochthonous reef occur above occur below the eastern half of the roof thrust. As noted the central anticline and lower, peripheral reef tract are above, in Barbados, the forearc basin strata of the roof serially younger, indicating the continuing rise of the central thrust's hanging wall are interpreted to be the Oceanic anticline/central diapir to the present. allochthon and syntectonic cover below the roof thrust is prism cover (shown as the Woodbourne Intermediate Unit INTERPRETATION OF THE in Fig. 10.10). GEOLOGY OF THE IFDB The movement history of the IFDB began in the early Miocene with an assumed forearc basin-prism configura- The geology and structural evolution of the IFDB are now tion31 (Fig. 10.10, top). The Atlantic lithosphere underrides interpreted using seismic data from Torrini and Speed31 the forearc basement just east (to the right) of the section. and the geology of Barbados. The principal rock units are: The prism detaches from, and is pushed west on, the base- Cenozoic forearc basin strata; accretionary prism rocks of ment, wedging below the forearc basin (Fig. 10.10, middle). which the basal complex of Barbados is assumed to be The wedging causes shortening in the forearc basin succes- representative; syndeformational cover, which is probably sion that is taken up by a long detachment fault that ends 50 like Miocene and younger prism cover on Barbados; and km west in a ramp and a fault-propagation fold, and by a diapirs5. The principal structures resolved or interpreted in train of detachment folds above the detachment. At this the IFDB are folds (Fig. 10.9), a long-floor detachment and stage, the forearc basin strata also move up the prism ramp a roof thrust (Fig. 10.10, bottom), and an extensive out-of- on the conjugate back thrust (Fig. 10.10, middle). Further sequence duplex between the floor and roof faults (Fig. westward motion of the prism is accommodated by ramping 10.10). The IFDB is divided into western and eastern zones within itself on one or more out-of-sequence thrusts. Ahead (Fig. 10.9) on the bases of the orientation and character of of it propagates a duplex of forearc basin strata below the folds at the seabed and by seismic properties. The western roof thrust (Fig. 10.10, bottom). zone is thought to be constituted by forearc basin strata in major folds in its western half and, in its eastern half, below

187 Barbados and the Antillean Forearc

Figure 10.9. Geologic map of inner forearc deformation belt (IFDB) showing eastern and western zones, and shallow structures (from Torrini and Speed ). Dotted contours are water depth in metres; T# is cross-section trace from Figure 10.10.

QUATERNARY REEFS AND THE RISE has the general form of a southwest-plunging arch, similar OF BARBADOS ISLAND to that of the topographic surface of the island (Fig. 10.11A), The limestone cap, together with the Oceanic allochthon, A large fraction of Barbados is covered by a limestone cap almost certainly once extended over what is now the Scot- (Fig. 10.5) that is comprised of Quaternary deposits related 4,6,13 land district (Fig. 10.5), from which they have been eroded. to coral reefs . The limestone cap is a relatively thin The coral reefs that compose the limestone cap occupy sheet, varying between about 30 and 120 m thick. The sheet adjacent strips within the cap (Fig. 10.11B) and are not

188 ROBERT C. SPEED

Figure 10.10. Cross-sections illustrating sequential development of IFDB by thin-skinned thrusting and arcward wedging of accretionary prism (from Torrini and Speed31).

vertically stacked one on the other. Each individual reef of and toward higher elevation4,6,1 3. For example, the ages of the more than 20 identified has a terrace shape in cross the crests of the lowest three terraces (Fig. 10.11B) are about section and three principal architectural elements (Fig. 80, 105 and 125 ka. The oldest of these, named First High 10.12), namely reef crest, backreef and forereef. The ele- Cliff (Fig. 10.11B), has been dated sufficiently thoroughly ments are analogous to biofacies of modern Caribbean to show that the duration of active coral growth on the crest reefs12 and have the reef crest defined by Acroporapalmata was at least 8 ka. The precision of dating decreases greatly that grew within a few metres of sea level. The forereef was with age, owing to many causes, but principally to the the seaward slope to the reef crest and is composed of debris isotopic exchange during carbonate diagenesis. The oldest from the crest together with deeper-water corals. The back- reef dated, near Golden Ridge (Fig. 10.11B), gives ages of reef was on the landward side of the crest and contains about 640 ka. The oldest preserved reefs are those rimming sediment deposited mainly in lagoons. the Scotland district; they could be as old as 1 Ma or more. The distribution of the crests of the major Quaternary It is evident that the succession of Quaternary reef tracts in reefs of the limestone cap is shown in Figure 10.12B13. the limestone cap record approximate sea level at various Individual crests can be traced by their A. palmata biofacies times in the past and that the pattern of outward migration from the northern to the southern end of the island. Collec- of sea level relative to the Scotland district records the tively, they form a series of parallel curvilinear tracks that progressive rise of Barbados island through sea level. The define the west-plunging arch; the Mount Hillaby-Clermont first point on Barbados to reach sea level was either in the nose (Fig. 10.11B) is the hinge of the arch and the locus of Scotland district or, perhaps, offshore to the east of the highest elevation of each reef crest. Scotland district. The distribution of reef crests in position Radiometric dating of corals of individual reef crests and elevation shows that the uplift has been heterogenous. indicates that, in general, age increases episodically from the The rates of uplift have been greatest in the Scotland district present shoreline toward the Scotland district (Fig. 10.11) and along the Mount Hillaby-Clermont nose (Fig. 10.11B).

189 Barbados and the Antillean Forearc

Figure 10.11. Maps of Barbados island showing features of Pleistocene reef cap (from Mesolella et al.13). (A) Elevation of island at base of reef obtained from well data. (B) Crests of individual reef tracts defined by Acropora palmata biofacies. Lined ornament indicates sub-reef outcrop.

-1 There, a rate of 0.4-0.5 m ka over 125 ka has been calcu- terrace and hence developed close to sea level. The implica- 4,11 lated . Elsewhere, rates are lower, approaching zero on tion is that the rise of Barbados island is due to local the south coast. tectonics. This rise can be explained as due to the However, the uplift of the island does not explain the upwelling of the mud-melange diapir of the central series of discrete reef terraces that make up the limestone anticline (Fig. 10.8). Evidence for this origin is that the cap, rather than a single continuous diachronous reef span- largest body of the diapir occurs below the highest ning the more-or-less million years since Barbados first elevations of the island as a huge, southwest-trending dyke reached sea level. The discrete reefs can be explained by and because diapirism is still active. However, the the superposition of eustatic oscillation of sea level on emplacement pattern of the diapir is complex. Smaller the presumably steadier tectonic rise of the island. The diapirs, which are probably satellite intrusions, cause local exposed reef terraces probably developed during peaks or deformation and the crest of the main body is flanged out high sea level stands of the oscillations, before they below the Oceanic allochthon (Fig. 10.8). emerged by tectonic uplift. The reefs that grew at other times in an oscillation were probably mainly covered by the next high stand growth. The 125 ka reef crest (First REFERENCES High Cliff; Fig. 10.11B) is correlated with the last interglacial period, when sea level was higher than that of 1 Argus, D. & Gordon, R.G. (in press). North America- today by about 6 m. The 80 and 105 ka reefs correspond to South America plate boundary. Journal of partial warming periods, called interstadials, in the late Geophysical Research. Pleistocene glacial epoch. 2Barker, L.H., Gordon, J.M. & Speed, R.C. 1988. A The tectonic uplift of Barbados island over the last study of the Barbados intermediate unit surface and million years is anomalous with respect to the subsidence of subsurface: in Barker, L.H. (ed.), Transactions of the much of the Barbados Ridge over the past 8 Ma. The Eleventh Caribbean Geological Conference, Dover unconformity occurs in the Barbados Ridge north and south Beach, Barbados, July 20th-26th, 1986, 36:1-36:24. of Barbados island31. This unconformity, now averaging 1.5 3Barker, L.H. & Poole, E. 1982. The Geology and Mineral km below sea level, is interpreted to have been a wavecut Resource Assessment of the Island of Barbados. Gov-

190 ROBERT C. SPEED

Figure 10.12. Diagrammatic cross-section through typical Barbadian reef tract showing principal morphologic and biofacial zones (from Mesolella12).

ernment of Barbados, St. Michael. Bulletin, 54,1899-1917. 4Bender, M.J., Fairbanks, R.G., Taylor, F.W., Matthews, 14Moore, J.C., Biju-Duval, B., Bergen, J.A., Blackington, RJC, Goddard, J.G. & Broecker, W.S/I979. Uranium- G., Claypool, G.E., Cowan, D.S., Duennebier, F., series dating of the Pleistocene reef tracts of Barbados. Guerra, R.T., Hemleben, C.H. J., Hussong, D., Marlow, Geological Society of America Bulletin, 90, 577-594. M.S., Nafland, J.H., Pudsey, C.J., Renz, G.W., Tardy, 5Brown,K.M.&Westbrook,G.K. 1989. Mud diapirism and M., Willis, M.E., Wilson, D. & Wright, A.A. 1982. subcretion in the Barbados Ridge accretionary wedge. Off scraping and underthrusting of sediment at the de- Tectonics, 7, 613-640. formation front of the Barbados Ridge: Deep Sea Drilling 6Edwards, R.L., Chen, J.H., Ku, T.L. & Wasserburg, G. Project Leg 78A. Geological Society of America 1987. Precise timing of the last interglacial period from Bulletin, 93,1065-1077. mass spectrometric determination of thorium 230 in 15Moore, J.C., Mascle, A., Taylor, E., Andreieff, P., Al- corals. Science, 236, 1547-1553. varez, F., Barnes, R., Beck, C., Behrmann, J., Blanc, G., 7Harrison, J.B. & Jukes-Brown, AJ. 1892. The Geology of Brown, K., Clark, M., Dolan, J., Fisher, A., Gieskes, J., Barbados. The Barbados Legislature. Hounslow, M., McLellan, P., Moran, K., Ogawa, Y., 8Larue, D.K. & Speed, R.C. 1983. Quartzose turbidites of Sakai, T., Schoonmaker, J., Vrolijk, P., WiUens, R. & the accretionary complex of Barbados, I, Chalky Mount Williams, C. 19&8. Tectonics and hydrogeology of the succession. Journal of Sedimentary Petrology, 53, northern Barbados Ridge: results from Ocean Drilling 1337-1352. Program Leg 110. Geological Society of America Bul- 9Larue, D.K. & Speed, R.C. 1984. Structure of the accre- letin, 100,1578-1593. tionary complex of Barbados, II, Bissex Hill. Geologi- 16Payne, P., Sargeant, M. & Jones, K. 1988. An approach to cal Society of America Bulletin, 95, 1360-1373. the evaluation and exploitation of hydrocarbons from 10Lawrence, S.R., Barker, L.H., Soulsby, A. & Payne, P. the Barbados accretionary prism: in Barker, L.H. (ed.), 1983. A geological and geophysical investigation of Transactions of the Eleventh Caribbean Geological Barbados. Abstracts, Tenth Caribbean Geological Conference, Dover Beach, Barbados, 20th-26th July, Conference, Cartagena, Colombia, 14th-20th August. 1986, 39:1-39:20. 11Matthews, R.K. 1972. Relative elevation of late Pleisto- 17Poole, E. & Barker, L.H. 1982. Geology of Barbados, cene high sea-level stands, Barbados. Quaternary Re- scale 1:50,000. Government of Barbados, St. Michael. search, 3,147-153. 18Saunders, J.B. & Cordey, W.G. 1968. The biostratigraphy of 12Mesolella, KJ. 1967. Zonation of uplifted Pleistocene the Oceanic Formation in the Bath Cliff section, coral reefs on Barbados, West Indies. Science, 156, Barbados: in Saunders, J.B. (ed.), Transactions of the 638-640. Fourth Caribbean Geological Conference, Port-of- 13Mesolella, K.J., Sealy, H.A. & Matthews, R.K. 1970. Spain, Trinidad, 28th March-12th April, 1965, 443- Facies geometries within Pleistocene reefs of Bar- 449. bados. American Association of Petroleum Geologists 19Saunders, J.B., Bernoulli, D., Muller-Merz, E., Oberhan-

191 Barbados and the Antillean Forearc

sli, H., Perch-Nielsen, K., Riedel, W.R., Sanfilippo, A. Hole, Massachusetts. & Tonini, R., Jr. 1984. Stratigraphy of the late middle 29Steineck, P.L, Breen, M, Nevis, N. & O'Hara, P. 1984. Eocene to early Oligocene in the Bath Cliff section, Middle Eocene and Oligocene deep seaOstracoda from Barbados, West Indies. Micropaleontology, 30, 390- the Oceanic Formation, Barbados. Journal of Paleon- 425. tology, 58, 1463-1496. 30 20Senn, A. 1940. Paleogene of Barbados and its bearing in Torrini, R, Jr. 1988. Structure and kinematics of the the history and structure of the Antillean-Caribbean Oceanic nappes, Barbados: in Barker, L.H. (ed.), region. A merican Association of Petroleum Geologists Transactions of the Eleventh Caribbean Bulletin, 24, 1548-1610. Geological Conference, Dover Beach, Barbados, 21Speed, R.C. 1981. Geology of Barbados: implications for 20th-26th July, 1986, 15:1-15:15. 31 an accretionary origin. Proceedings of the 26th Inter- Torrini, R, Jr. & Speed, R.C. 1989. Tectonic wedging in national Geological Congress, 259-265. the forearc basin-accretionary prism transition, Lesser 22Speed, R.C. 1983. Structure of the accretionary complex Antilles forearc. Journal of Geophysical Research, 94, of Barbados, 1, Chalky Mount. Geological Society of 10549-10584. 32 America Bulletin, 94, 3633-3643. Torrini,R, Jr., Speed, R.C. &Claypool,G.E. 1990. Origin 23Speed, R.C. 1988. Geological history of Barbados: a and geologic implications of the diagenetic limestone preliminary synthesis: in Barker, L.H. (ed.), Transac- in fault zones of Barbados: in Larue, D.K. & Draper, G. tions of the Eleventh Caribbean Geological Confer- (eds), Transactions of the Twelth Caribbean Geologi- ence, Dover Beach, Barbados, 20th-26th July, 1986, cal Conference, St. Croix, U.S. Virgin Islands, 7th-llth 29:1-29:11. August, 1989, 366-370. 33 24Speed, R.C. 1990. Volume loss and defluidization history Torrini, R., Jr., Speed, R.C. & Mattioli, G.S. 1985. Tec- of Barbados. Journal of Geophysical Research, 95, tonic relations between forearc basin strata and the 8983-8996. accretionary complex at Bath, Barbados. Geological 25Speed, R.C. & Larue, D.K. 1982. Architecture and impli- Society of America Bulletin, 96, 861-874. 34 cations for accretion. Journal of Geophysical Research, Westbrook, G.K. 1975. The structure of the crust and 87,3633-3643. upper mantle in the region of Barbados and the Lesser 26Speed, R.C., Torrini, R., Jr. & Smith, P.L. 1989. Tectonic Antilles. GeophysicalJoumal of the Royal Astronomi- evolution of the Tobago Trough forearc basin. Journal cal Society, 43, 201-242. 35 of Geophysical Research, 94, 2913-2936. Westbrook, G.K. 1982. The Barbados Ridge complex: 27Speed, R.C. & Walker, J.A. 1991. Oceanic crust of the tectonics of a mature forearc system: in Leggett, J.K. Grenada Basin in the southern Lesser Antilles Arc (ed.), Trench and Forearc Geology. Special Publica- Platform. Journal of Geophysical Research, 96, tion of the Geological Society, London, 10, 275-291. 36 3835-3851. Westbrook, G.K., Ladd, J.W., Buhl, P., Bangs, N. & Tiley, 28Speed, R.C., Westbrook, O.K., Ladd, J.W., Mauffiet, A., G.J. 1988. Cross section of an accretionary wedge: Biju-Duval, B., Mascle, A., Jackson, R., Munschy, M, Barbados Ridge complex. Geology, 16, 631-635. 37 Stein, S., McCann, W.R., Schoonmaker, J.E., Saunders, Westbrook, G.K., Smith, M.J., Peacock, J.H. & Poulter, J.B., Beck, C. & Wright, A. 1984. Lesser Antilles arc M.J. 1982. Extensive underthrusting of undeformed and adjacent terranes, Atlas 10, Ocean Margin Drill- sediment beneath the accretionary complex of the ing Program, Regional Atlas Series. 27 sheets, scale Lesser Antilles subduction zone. Nature, 300, 625- 1:2,000,000. Marine Science International, Woods 628.

CHAPTER 11

Tobago

TREVOR A. JACKSON and STEPHEN K. DONOVAN

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION showing the distribution of these different rock types. Fol- lowing this initial contribution, Trechmann34 published an THE ISLAND of Tobago is located in the southeastern account of the Cenozoic sedimentary rocks and palaeontol- corner of the Caribbean Plate and has been interpreted as ogy of the island. part of an allochthonous terrane that forms the easternmost 29 It was not until John Maxwell of Princeton University fragment of the Caribbean Mountain System . The island visited the island that Tobago was mapped in detail. Max- can be divided into essentially three provinces; one sedi- 15 well's research was for a doctoral degree and the bulk of his mentary, one igneous and one metamorphic . Snoke and thesis was eventually published in the Bulletin of the Geo- Rowe divided these provinces into five major units: (1) logical Society of America20. His map (Fig. 11.1 herein) and North Coast Schist Group (NCSG); (2) Tobago Volcanic stratigraphy (Table 11.1 A) of the island have been generally Group (TVG); (3) ultramafic -felsic plutonic suite; (4) mafic accepted, with only minor modifications, by subsequent dyke swarm; and (5) Tertiary and Quaternary sedimentary workers. Indeed, in the revised geologic map (Fig. 11.2) of rocks. the island by Snoke et al.29, the names of groups and several The metamorphic province contains the oldest of these formations have been retained (Table 11.IB). Maxwell's units, the North Coast Schist Group, comprising a succes- work has therefore formed the basis on which the past 50 sion of Lower Cretaceous(?) low-to-medium grade meta- years of research on the island have been developed. morphic rocks that outcrop in the northern third of the island. Maxwell20 (Table 11.lA herein) considered the NCSG The NCSG is juxtaposed to rocks of the igneous province, to represent the oldest rocks exposed on Tobago, compris- which consists of a complex pluton of batholithic propor- ing, in order of descreasing age, the Main Ridge, Parlatuvier tions that intruded the Tobago Volcanic Group, a thick and Mount Dillon Formations. The mineralogy of these sequence of volcaniclastic rocks and related lava flows of formations shows them to be low-to-medium grade, region- maf ic to intermediate composition. Mafic dykes intrude the 29 ally metamorphosed rocks. The Main Ridge Formation pluton, TVG and NCSG . The sedimentary province is comprises phyllitic sericitic schists which grade upwards located in the southwestern end of the island and comprises into the series of metavolcanic rocks known as the Parlatu- Neogene and Quaternary terrigenous and carbonate rocks. vier Formation. Maxwell20 considered that "Mica schist, These rocks lie unconformably on the pre-Cenozoic rocks thin beds of quartzite and chert, and interbedded metavol- of the igneous province. canics characterize the Mount Dillon Formation." To the south and southeast of the NCSG, Maxwell identified younger plutonic and volcanic rocks. He consid- PREVIOUS WORK ered the ultramafic plutonic series to be the oldest unit of The earliest general account of the geology of Tobago was written by Cunningham-Craig5. This report described the Figure 11.1. (next page) Geological map of Tobago (after occurrence of schists, igneous rocks, Tertiary beds and Maxwell20). Maxwell's stratigraphy is summarised in coral, and included a geological sketch map of the island Table 11.1A.

193

194 T. A. JACKSON and S. K. DONOVAN

tbis sequence, followed by extrusive and intrusive rocks of Rowley and Roobol25 also produced several K-Ar ra- the Tobago Volcanic Group, and by a complex dioritic diometric dates for both the TVG and the plutonic rocks. intrusion of batholithic proportions. Unfortunately, several of the isotopic dates did not agree Maxwell divided the TVG into four formations. The well with the field evidence and it was concluded that Goldsborough (oldest) and Hawk's Bill (youngest) Forma- metasomatism might have been responsible for producing tions were both described as dacitic, with the Hawk's Bill the anomalous ages. This phenomenon was confirmed sub- consisting mainly of lava flows, and the Goldsborough sequently by Jackson and Smith16 as the major cause of rock comprising tuffs, breccias and flows. Both formations were alteration in the TVG and it was postulated that metasoma- delineated as occupying small coastal areas of outcrop rela- tism most probably occurred during the intrusion of the tive to the more extensively exposed Bacolet and Merchis- batholith. ton Formations. Maxwell reported that the Bacolet and Additional K-Ar dates were supplied by Girard10 who, Merchiston Formations are petrographicaUy identical, and by combining her results with some of the more 'acceptable4 consist of basalt and andesite agglomerates, tuff-breccias results from Rowley and Roobol25, narrowed the timing of and tuffs. the igneous activity in Tobago to between 100 and 70 Ma. The diorite batholith was interpreted by Maxwell to be These ages concurred with the palaeomagnetic data publish- younger than the ultramafic rocks. This relationship has ed by Briden et al4, who reported that the igneous rocks of been disproven by Snoke et al.29, who have shown that the Tobago all display a normal polarity which is consistent ultramafic rocks formed from cumulates associated with the with the long Cretaceous normal polarity interval from circa crystal differentiation of the 'so-called' diorite batholith. 110 to 80 Ma. More recently, 40Ar-39Ar dates reported by However, the heterogenous nature of the batholith was Sharp and Snoke27 provided a range for the volcanic-plu- noted by Maxwell, who identified at least three intrusive tonic-dyke complex of between 108 ± 3 to 91 ± 2 Ma which phases: melagabbro and meladiorite; mesocratic hornblende is in agreement with previous published ages. diorite; and biotite quartz diorite. Despite the post-magmatic alteration of the volcanic Maxwell reported that the Cretaceous igneous rocks rocks, Girard10 and Leterrier et al18 were able to determine were overlain unconformably by late Tertiary and Quater- the composition of some of the unaltered pyroxene and nary fossiliferous sedimentary rocks in the south. The Ter- amphibole phenocrysts in the TVG using microprobe analy- tiary rocks of the Rockly Bay Formation occur as a series of sis. The authors established that the augite of the Bacolet small, scattered exposures and comprise a sequence of well- Formation was characteristic of pyroxenes associated with bedded marls, sands and sandy clays. Maxwell considered calc -alkaline rocks, whereas the diopsidic augite identified this formation to be late Miocene to Pliocene in age. How- in the Hawk's Bill Formation commonly occurs in tholeiitic ever, Saunders and Muller-Merz26 dated this formation as rocks. They also identified magnesio-hastingsite amphi- late early to early middle Pliocene in age on the basis of boles which are associated with arc-related rocks. foraminiferal data. Jackson et al.15 also established an arc-related origin The Quaternary beds, which are primarily coralline for the older metavolcanic rocks of the NCSG. Geochemical limestone, are far more extensive in areal extent than the analyses of the Parlatuvier Formation showed that prior to Tertiary rocks. According to Maxwell, these rocks outcrop undergoing low-grade metamorphism, the rocks were oce- over much of southwest Tobago, at elevations of up to 30m. anic island arc basaltic and andesitic lava flows and volcani- Following Maxwell's major contribution, there was clastics, thus belonging to the primitive island arc or island very little active research on the geology of Tobago for arc tholeiite series. Subsequently, Frost and Snoke9 also almost 30 years. The resurgence of interest began with a determined a volcanogenic source for the Mount Dillon paper by Rowley and Roobol25, which included geochemi- Formation, which they described as silicified tuffs and cal analyses for the TVG and the ultramafic -diorite intru- cherts. sion. They showed that chemical similarities exist between the Hawk's Bill and Goldsborough, and Bacolet and Mer- chiston, Formations, thus supporting the petrographic simi- STRATIGRAPHY larities previously detected by Maxwell. Donnelly and Rogers7 published further chemical analyses, confirming Mesozoic that the volcanic and plutonic rocks of Tobago were charac- The Mesozoic rocks of Tobago are divided into three teristic of the primitive island arc series. Thus, for the first major groups: the greenschist facies NCSG; the TVG; and time, a volcanic arc setting associated with plate conver- the ultramafic -felsic plutonic suite. A mafic dyke swarm gence was established for the TVG and the consanguineous intruded the TVG and the plutonic suite. Scattered dykes ultramafic -diorite batholith. have also been identified in the NCSG, but are much rarer.

195 Tobago Table 11.1A. Stratigraphy of Tobago (after Maxwell20).

SYSTEM SERIES FORMATION Quaternary Coral ------Tertiary Pliocene and Upper Miocene Rockly Bay Formation ------UNCOMFORMITY------Hawk's Bill Formation ------TOBAGO Intrusion of diorite batholith VOLCANIC ------GROUP Merchiston-Bacolet Formation ------Goldsborough Formation Cretaceous(?) ------UNCOMFORMITY------Intrusion of ultramafic rocks NORTH ------COAST Mount Dillon Formation SCHIST ------GROUP Parlatuvier Formation ------Main Ridge Formation

In this section, we have chosen to use, where possible, The Parlatuvier Formation weathers to a yellowish-brown or the revised stratigraphy and lithological subdivisions of buff colour, similar to fresh and weathered Mount Dillon Tobago defined by Frost and Snoke9 and Snoke et al.29 Formation. However, fresh exposures of the Parlatuvier Formation differ from the Mount Dillon Formation by being North Coast Schist Group (NCSG) pale green in colour because of the abundance of chlorite, Frost and Snoke 9 and Snoke et al29 divided the NCSG epidote and amphibole. The Parlatuvier Formation contains into two lithostratigraphic units, the Parlatuvier Formation porphyroclasts of plagioclase feldspar, amphibole and cli- and the structurally higher Mt. Dillon Formation. The main nopyroxene enclosed in a fine-grained schistose matrix of difference between this revised stratigraphy and that of mica, chlorite, granular quartz and epidote, and actinolite. Maxwell is the suppression of the Main Ridge Formation Jackson et al.15 and Snoke et al.29 have postulated that (which is regarded as part of the Parlatuvier) and an associ- prior to pervasive lower greenschist facies metamorphism, ated argillite-volcaniclastic interval within the Parlatuvier the Parlatuvier Formation comprised basalt and andesite (compare Figs 11.1 and 11.2A, and Tables 11.1A and lava flows, lapilli tuffs, crystal lithic tuffs and mafic dykes. 11.1B). On the basis of the geochemical evidence (in particular, the immobile and discriminant elements), it was concluded that Parlatuvier Formation the metavolcanic rocks of the Parlatuvier Formation were The Parlatuvier Formation is more extensive than the originally extruded in an oceanic island-arc setting. Normal- Mount Dillon Formation, which outcrops only along the ized rare earth element (REE) patterns ar e either flat or show southwestern edge of the metamorphic belt (Fig. 11.2A, B). slight LREE enrichment, suggesting that these rocks belong

196 T. A. JACKSON and S. K. DONOVAN

Table 11.1B. Stratigraphy of Tobago (slightly modified after Frost and Snoke9).

SYSTEM SERIES FORMATION Holocene Superficial deposits ------Quaternary Coral limestone Pleistocene ------Montgomery beds Tertiary Pliocene ------Rockly Bay Formation ------UNCOMFORMITY------Mafic dyke swarms, ultramafic-felsic plutonic intrusions ----INTRUSIVE CONTACT---- TOBAGO Bacolet Formation ‘medial’ VOLCANIC ------GROUP Epiclastic Unit Cretaceous ------Goldsborough Formation ------Argyle Formation ------UNCOMFORMITY------Mount Dillon Formation Lower NCSG ------Parlatuvier Formation

to the primitive island arc/island arc tholeiite series15. genic argillites. There is evidence from both the chemistry and petrography that suggests many of the metavolcanic Mount Dillon Formation layers were affected by siliceous hydrothermal alteration The Mount Dillon Formation consists of a greenschist prior to metamorphism9. facies sequence of interbedded mica schists, quartzites and phyllites. together with minor graphitic schists and chert bands1,20. Schists are the dominant lithology within this Figure 11.2A (next page). Geological map of Tobago formation. They are fine-grained and contain relict porphy- (reproduced by permission of the Geological Society roclasts of plagioclase feldspar and quartz in a matrix of from "Tobago, West Indies, a fragment of a Mes ozoic chlorite, mica, stilpnomelane, quartz and opaques1,30. island arc: petrochemical evidence” by C.D. Frost and Snoke et al.30 suggested a volcanogenic source for A.W. Snoke, in Journal of the Geological Society, much of the Mount Dillon Formation, with the rocks being London, volume 146 for 1989). See also Snoke et predominantly meta-dacite crystal-lapilli tuffs and volcano- al.29,30 and Table 11.1B herein.

197

198

199 Tobago

Figure 11.3. Simplified geological map of southwest Tobago (after Donovan8, redrawn after Saunders and Muller-Merz26). The Montgomery beds and Rockly Bay Formation are not differentiated from each other in this figure. The sections in Figure 11.4 were measured at the southwestern end of the outcrop in Rockly Bay. The inset map shows the position of Tobago (arrowed) in the eastern Caribbean. Key: lrb=Little Rockly Bay.

Amphibolitic rocks relative to the NCSG (G. Wadge, written communication). Yule et al.36 identified an east-west striking belt of The normal fault that defines the southern boundary of the amphibolitic rock, less than 250 m thick, along the northern aureole is younger and may have been produced when margin of the ultramafic-felsic batholith. This amphibolitic strike-slip movement began along the southern margin of rock possesses a weak to moderate foliation that overprints the Caribbean Plate during the Neogene. the pre-existing fabric of the NCSG wallrock, and contains mainly hornblende and plagioclase ± epidote. Snoke et al30 Tobago Volcanic Group interpreted these rocks as a metamorphosed aureole of am- The contact between the TVG and the NCSG is inferred phibolite facies in the NCSG wallrock, produced during the to be an unconformity, the TVG post-dating the deformation emplacement of the ultramafic-diorite batholith. and greenschist metamorphism of the NCSG. Yule et al.36 defined the northern boundary of the The revised stratigraphy of the TVG by Frost and amphibolitic rocks on the presence of a high-angle thrust Snoke9 is incomplete and the authors admit to some uncer - fault, while the southern boundary is marked by a normal tainty in the stratigraphic order of the proposed formations fault. Snoke et al.30 proposed that after the initial contact with the inclusion of 'undifferentiated volcanic and sedi- metamorphism of the NCSG wall rock, the plutonic commentary rocks' (Fig. 11.2). The TVG mainly comprises plex and its associated aureole rose to cooler levels of the clinopyroxene-plagioclase phyric basalt and andesite lava 20 crust. During this process, a distinct brittle reverse fault was flows, volcaniclastic breccias and tuffs , with minor beds created (the 'back aureole' fault of Snoke et al.). Subsequent of volcanogemc argillites bearing radiolanans and mouldic 28 laboratory analysis of oriented samples has shown that the ammonites . The radiolarians indicate an Albian to early principal deformation was shear with extension, indicating Cenomanian age, while the ammonites suggest these beds 9,30 that the sense of movement was normal and probably cre- to be of Albian antiquity . These fossiliferous layers, ated by downward movement or collapse of the pluton which occur in association with volcaniclastic breccia, grit

200 T. A. JACKSON and S. K. DONOVAN

and sandstone (Table 11.IB) have been mapped as a sepa- The outermost unit is the deformed mafic plutonic -vol- rate 'Epiclastic Unit' by Frost and Snoke9. Frost and Snoke9 canic complex, which outcrops along the northern margin also introduced the Argyle Formation, which has an abun- of the batholith. These rocks were penetratively deformed dance of hornblende phenocrysts in the lithic fragments of under amphibolite to pyroxene hornfels facies during the the breccia and tuff. Hornblende is generally rare or absent emplacement of the adjacent ultramafic unit of the pluton31. in the Bacolet and Goldsborough Formations, where plagio- The common minerals of the ultramafic rocks are cli- clase and clinopyroxene are the two dominant phenocrystic nopyroxene and olivine±hornblende and spinel. This unit minerals. Indeed, the proportions of these two minerals are outcrops as a linear belt that ranges in composition from a major factor in differentiating between the Bacolet and wehrlite to olivine clinopyroxenit^ with occasional dunite Goldsborough Formations; Goldsborough has plagio- masses 29. These rocks are regarded as the early cumulate clase>clinopyroxene, whereas the Bacolet has clinopy- facies of the plutonic suite24. roxene>plagioclase (although Wadge and co-workers do The main phase of plutonic activity resulted in the not consider that these non-Bacolet Formation plagioclase- intrusion of the heterogenous diorite/gabbro unit. These rich rocks constitute a true formation; written communica- rocks are mainly medium- to coarse-grained, but are occa- tion). sionally pegmatitic. Hornblende is the most abundant ferro- Although dominantly a volcaniclastic succession, the magnesian mineral and can comprise as much as 75% of the TVG also includes minor lava flows and pillowed lava rock volume20. Clinopyroxene and plagioclase are also ma- flows. The most impressive pillow lavas outcrop between jor rock-forming minerals in the plutonic rocks, which range Mount Irvine and Plymouth (Fig. 11.2), and represent a in composition from melagabbro/meladionte to quartz dio- major portion of what Maxwell described as the Hawk's Bill rite. Formation. The rocks contain amygdaloidal quartz and The biotite tonalite (=biotite quartz diorite of Max- veins of red jasper, suggesting that siliceous hydrothermal well20) forms an east-west dyke-like body (Fig. 11.2) and is fluids played an active, post-magmatic role in the alteration presumed to be the youngest part of the plutonic suite31. of these lavas. Because of the high silica content of these This unit contains biotite, hornblende, oscillatory zoned rocks, Maxwell20 erroneously described them as dacites. plagioclase, quartz and minor opaques, forming the most Subsequently, Jackson and Smith16 and Frost and Snoke9 siliceous unit in the plutonic complex9. Small sections of the have shown that the rocks range from basalt to andesite in tonalite, like the other units of the pluton, have undergone composition. deuteric alteration, with biotite altered to chlorite and leu- Metasomatism has been responsible for the extensive coxene, hornblende changed to chlorite and epidote, and alteration of the TVG, based on the mobility of K, Rb and plagioclase altered to sericite and clay minerals. Ba in these rocks16, and the occurrence of pumpellyite, albite, zeolite, sericite, chlorite and serpentine. Mafic volcanic dyke complex 20 The TVG characteristically has a primitive island arc Maxwell identified volcanic dyke rocks in the plu- chemistry6,7. The primitive island arc nature of these rocks tonic suite, the TVG and the NCSG. The dykes are generally is particularly well displayed in their relatively flat REE fine-grained, containing hornblende ± clinopyroxene and 9 patterns and their high FeO/MgO to silica ratio. Radiogenic plagioclase as the major constituents. Frost and Snoke Sr and Nd isotopic data9 suggest that the TVG formed in an noted that the dykes that intrude the NCSG are chemically oceanic island-arc setting that was far removed from conti- distinct from those that occur in the younger igneous rocks. nental influences. They consider the dykes that intrude the NCSG to be andesite, containing low MgO (< 4%) and FeO (5-6%), and Ultramafic-felsic plutonic suite possessing a characteristic greenschist facies mineral as- The ultramafic -felsic plutonic complex intrudes both semblage. In contrast, the dykes which occur in the TVG the NCSG and the TVG, and forms a zone of contact and the plutonic complex range from basalt to andesite, and metamorphism in those areas that have been unaffected by have higher percentages of FeO (>7%) and MgO (>8%). faulting. Rocks of the TVG are occasionally found incorporated within the pluton as large-scale roof pendants28. Cenozoic Snoke et al.31 have shown the plutonic suite to be The Cenozoic rocks of southwestern Tobago are Neo- composite, consisting of four distinct and mappable units: gene in age and late Quaternary in age and, apart from deformed and metamorphosed mafic rocks; ultramafic superficial deposits, originated in shallow marine environ- rocks; diorite-gabbro; and biotite tonalite. These divisions ments. These rocks rest unconformably on the underlying are different to those originally identified by Maxwell20 (see Mesozoic strata. The Cenozoic sequence is flat-lying, struc - 31 above). turally simple apart from rare faults and has been divided

201 Tobago

Figure 11.4. Strip logs (after Donovan8) of two measured sections at the southwestern end of the type section of the Rockly Bay Formation, drawn from data in Trechmann34 and Maxwell20. Note the apparently large variation in thickness of some beds, in two sections measured within 100m of each other. into four units (Fig. 11.2): the Rockly Bay Formation of of clays and coarser clastic rocks, including pebble beds, Maxwell20; the Montgomery beds of Snoke and co-workers; derived from weathering of Mesozoic (and Eocene; see the Pleistocene raised reef; and superficial deposits (river above) rocks and incorporates an abundant marine fauna alluvium and beach sands) of late Quaternary age. (marl beds with Ostrea sp. and balanids should probably be included within this formation3,34). The type section is Paleogene exposed along the coast at Rockly Bay, southwest of Scar - Reworked Eocene orbitoid foraminifers at the base of borough and towards Red Point (Fig. 11.3). Both the Neogene Balanus Clay Bed of the Rockly Bay Forma- Trechmann34 and Maxwell20 published measured sections of tion (Renz in Maxwell20) provide the only evidence for the type sequence, which are summarized in Figure 11.4. Paleogene deposition in the Tobago area. Borehole evidence35 suggests a maximum thickness of circa 180 m for this formation, beneath a cover of coral Rockly Bay Formation limestone, with sands and pebble beds in the lower most The Rockly Bay Formation3,20 comprises a sequence third overlain by clay-rich strata. 202 T. A. JACKSON and S. K. DONOVAN

Figure 11.5. Schematic reconstruction of the palaeoenvironment of the mid-Pliocene Balanus Bed of Tobago (after Donovan8). The full vertical development of the bed is not shown. Relationships between localities not to scale. Vertical scale approximates both to water depth and bed thickness.

Trechmann34 considered the Rockly Bay Formation Plio-Pleistocene in age. (=Tobago "Crag" of Trechmann) to be Miocene in age on Trechmann34 recognised two horizons with distinctive the basis of its mollusc fauna (Table 11.2). Maxwell20 macrofossil assemblages in the Rockly Bay Formation. The suggested it to be comparable to the lower part of the Talparo Arca patricia Bed of Little Rockly Bay (Fig. 11.3) overlies Formation of Trinidad (latest Miocene to Pliocene in age), the unconformity with the Mesozoic volcanic rocks. The based on a comparative analysis of the foraminifers by H.H. bed includes disarticulated valves of the large ark Anadara Renz. Saunders and Muller-Merz26 showed the Rockly Bay (='Arca’) patricia and oysters referred to Ostrea sp. A basal Formation to be late early or early middle Pliocene in age conglomerate is well-developed and occasional pebble beds following identification of a diverse fauna of foraminifers. occur throughout this part of the sequence (see measured On the basis of this age, Donovan8 speculated that the sections in Trechmann34 and Maxwell20). deposition of the Rockly Bay Formation may have been The second distinct fossiliferous horizon is the Balanus related to a mid-Pliocene rise in sea-level of about 100 m11. Clay Bed (Maxwell20; forming part of the Tobago "Crag" of However, Mr. R.D. Liska (written communication, May Trechmann34; Fig. 11.4 herein). The invertebrate macro- 3rd, 1990) has subsequently informed SKD that a restricted fauna is dominated by the large balanid barnacle and well-defined palynoflora of Pleistocene antiquity has Megabalamis tintinnabulum (Linne) sensu lato (a faunal list been recovered from this unit. The Miocene age of is given in Table 11.2). In the type section of the Rockly Bay Trechmann is discounted due to the inherent difficulties of Formation, the matrix of the Balanus Clay is muddy, with using Neogene molluscs from the western Atlantic for such about 60% or more by weight of the sediment dominated by 33 Lyellian determinations . Until the conflicting micropa- balanid shells and disarticulated plates. This is an unusual laeontological determinations are resolved, it is considered environment for balanids, which prefer a hard substrate, but parsimonious to regard the Rockly Bay Formation as being which occur as fist-size aggregations growing from an origi-

203 Tobago

nally small attachment point. It is also unusual for barnacles to form such a dominant element within a fauna. Laterally, the muds grade into sands which are more proximal to the Table 11.2. Macrofauna of the Balanus Clay weathering igneous source. Indeed, close to the lateral un- Bed, Rockly Bay Formation. Molluscan names 34 conformable contact with the Mesozoic volcanic rocks, in from Trechmann included without revision. situ weathered basalt is superficially identical with sands 8 enclosing barnacles, oysters and scallops . The Balanus Barnacles (Withers in Trechmann34) 34 Clay Bed presumably overlies or overlaps the Arca patri- Megabalanus tintinnabulum (Linnaeus) cia Bed, and is in turn overlain by clays and fine sands in sensu lato Balanus amphitrite? Darwin 26 which macrofossils are rare . 20,34 Donovan8 (Fig. 11.5 herein) has modelled the pa- Bivalves laeoenvironment of the Balanus Clay Bed. The apparently Glycimeris cf. acuticostatus Carrick Moore wedge-shaped morphology of this unit was probably the Pecten cf. maturensis Maury result of a rise in sea level with associated transgression. The Spondylus cf. bostrychites Guppy lateral contact with the basalt is a rare example of an ancient Ostrea sp. 17 rocky shoreline being preserved in the rock record and Gastropods20,34 here, at least, suitable substrates for barnacle attachment Conus cf.forvoides Gabb were plentiful, so individuals did not aggregate as they did Ancilla (Eburnd) sp. in the clay environment of the type section. Grain size and Malea cf. elliptica Pilsbry and Johnson bed thickness diminish away from this contact. Natica cf. catena Da Costa Montgomery beds Scalaria sp. 29,31 Fasciolaria semistriata G.B. Sowerby Snoke et al. (Fig. 11.2A herein) have separated a Fasciolaria sp. sequence of "sandstone, conglomerate, and limestone of Fissurella sp. Montgomery" from the Rockly Bay Formation. These Vermetus sp. Montgomery beds are awaiting formal description, but are postulated to be of Plio-Pleistocene age and are inferred to Echinoids19 be younger than the Rockly Bay Formation. Eucidaris tribuloiaes (Lamarck) Arbacia improcera (Conrad) Coral limestone The youngest lithified deposit in southwest Tobago is Bryozoans (Dr. P.D. Taylor, research in progress) the coral limestone, a Pleistocene raised reef of probable last circa 10 genera 35 35 interglacial age . Wadge and Hudson estimated that this Crabs deposit has an average thickness of 12 m over an area of 27 34 decapod fragments indet. sqr km. Trechmann published a brief faunal list for the coral limestone, although it is undoubtedly incomplete; the dominant faunal elements are benthic molluscs and her - matypic corals. Both Trechmann34 and Maxwell20 published of beach sands in Tobago is strongly influenced by weath- measured sections from this sequence. Unfortunately, a ering and erosional products derived from the Mesozoic detailed palaeoenvironmental analysis of this late Pleisto- igneous and metamorphic sequences of the island. However, where protected from a terrigenous input, carbonate beach cene raised reef, perhaps comparing it with the modem 13 Buccoo Reef12 of southwest Tobago, has yet to be under- sands have developed . taken (but see Snoke et al.30). Wadge and Hudson35 dis - agreed with Trechmann's assertion34 that the coral limestone is terraced, although Maxwell (in Beckmann et STRUCTURE al.3) considered that there is a "...series of ill-defined terraces from the lowest at about six feet [1.85 m] to remnants Deformational features of the metamorphic rocks at about 100 feet [circa 30 m] above sea level." The NCSG is comprised of strongly deformed metavolcanic and metasedimentary rocks which experienced pene- Superficial deposits trative plastic deformation during greenschist facies Research on the superficial deposits of Tobago metamorphism. Ahmad et al.1 recognised two phases of 2,13,14,22 has concentrated on the beach sediments . Composition deformation within the NCSG, characterised by synmetamorphic ductile deformation (D1) and superposed brittle

204 T. A. JACKSON and S. K. DONOVAN

23 deformation (D2) features. According to Rowe , both linear The youngest faults occur in the south of Tobago within (L) and planar (S) fabrics were developed during meta- the sedimentary province and are associated with the E-W 21 morphism, producing a series of L- and S-tectonites. striking Southern Tobago Fault System. Morgan et al. Non-penetrative brittle structures (such as upright, kink- reported that this fault system is active and has a right lateral style folds and crenulations) and brittle fault zones cross-cut motion based on the earthquake swarm data for 1982. This and deform the earlier ductile fabrics. motion conforms with a broad zone of right-lateral plate The first generation D1 structures were formed during boundary motion displayed by other active faults along the the deformation and metamorphism of the NCSG. They are southeastern corner of the Caribbean Plate. represented by F1 isoclinal folds that developed around a low, plunging NE-S W fold axis and by a pervasive regional GEOLOGICAL HISTORY foliation, S1, that is subparallel to the axial surfaces of the 32 23 F1 folds . Rowe stated that "S1 foliation varies in character from weakly developed, closely spaced cleavage in LS The Mesozoic metamorphic and igneous rocks that form the tectonites to strongly developed slaty and phyllitic cleavage greater part of Tobago represent two stages of oceanic arc in SL tectonites." growth with an intervening phase of regional metamorphism. Prior to deformation and metamorphism of the D2 structures, which represent a second episode of NCSG, basalt, andesite and dacite volcaniclastic rocks and deformation, are characterised by F2 folds and a local spaced 1 minor lava flows of primitive island arc/island arc tholeiite to crenulation cleavage S2 . The F2 folds are tight to open affinity were extruded on a basement of oceanic crust. This and have developed approximately coaxial to F1 fold axes. first stage of arc development occurred in either the late The associated planar fabrics are only sparsely developed 9 and are best observed within the incompetent layers of the Jurassic or early Cretaceous . Mount Dillon Formation1. The second stage of arc growth occurred in the mid Structural elements also occur in the amphibolite facies Cretaceous, and was preceded by penetrative deformation aureole and include synmetamorphic foliations and mineral and lower greenschist facies metamorphism of the protolith. 32 This younger arc was built directly on, or adjacent to, the lineations, and a post-metamorphic open fold phase . 9 These structures are best displayed adjacent to the ul- older arc terrane represented by the NCSG . The second tramafic pluton where strong L-S tectonites are developed, stage was manifested by subaqueous primitive island arc/is - with foliation striking subparallel to and dipping beneath the land arc tholeiite basalt and andesite volcanic breccias, tuffs plutonic complex, and the lineation plunging gently to mod- and lava flows that form the TVG, intruded by an associated erately down the dip of the foliation23. The foliations are suite of ultramafic -tonalite plutons. The emplacement of the defined by the alignment of the hornblende and plagioclase plutons led to the alteration of the TVG and the NCSG, crystals, and by the alternating hornblende- and plagioclase- producing a strongly foliated and lineated amphibolite fa- rich gneissic layers. Lineations are defined by the elongate cies aureole. The last phase of magmatic activity during the hornblende and plagioclase grains 23. growth of the mid-Cretaceous volcanic-plutonic arc complex was the intrusion of a mafic dyke swarm. Rowe23 Fault systems considered that the preferred orientation of the dykes and There are several major faults that transect the lithic the presence of small-scale extensional faulting suggest that belts of Tobago. The two prominent trends of these major the arc may have evolved in an extensional, rather than fault systems are E-W and NNW-SSE. These faults occur contractional, setting, thus facilitating the emplacement of mainly in the igneous and metamorphic terranes, with the the pluton to a higher level within the crust. oldest being the ‘back aureole' fault which defines the Uplift and erosion of the composite arc was accompa- northern contact of the amphibolite facies aureole (see nied by normal and oblique-slip faulting during the late above). It is a normal fault whose origin is probably linked Mesozoic and Paleogene. Possible downfaulting in the to the downward movement or collapse of the pluton relative south was coupled with a mid-Pliocene(?) rise in sea level to the NCSG (G. Wadge, written communication). The which led to the deposition of the Rockly Bay Formation. 'back aureole' fault is offset by younger normal and oblique- During the Pleistocene there was normal, followed by re- slip faults32. The normal faults strike approximately E-W verse, movement along the Southern Tobago Fault System. and are subparallel to the ‘back aureole' fault. Both sets of This movement accounted for the deposition, and sub- 35 faults are offset by NNW-SSE striking, high-angle, oblique- sequent uplift and tilting, of the coral limestone . slip faults (Fig. 11.2). According to Snoke et al.32, the age ACKNOWLEDGMENTS—The authors thank the Geological Society, of the 'back aureole' fault is Cretaceous, but the normal and London, and Dr. A.W. Snoke for granting permission to reproduce Figure oblique-slip faults are Tertiary. 11.2A, and the Geological Society of Trinidad and Tobago and Dr. A.W. 205 Tobago

Snoke for granting permission to reproduce Figure 11.2B. Fieldwork by London, 146, 953-964. SKD in Tobago was supported by the 1987 Sylvester-Bradley Award of 10 the Palaeontological Association, the Geologists' Association (G.W. Girard, D. 1981. Petrologie du quelques series spilitiques Young Fund and G.A. Fund), and the Geological Society of Trinidad and Mesozoiques du domaine caraibe et ensembles Tobago. TAJ gratefully acknowledges support from the Ministry of Energy magma-tiques de Vile de Tobago: implications geody- and Natural Resources, Government of Trinidad and Tobago, for logistic namiques. Unpublished Ph.D. thesis, Universite de support during fieldwork in 1982 and 1984. Bretagne Occidentale. 11Haq, B.U., Hardenbol, J. & Vail, P.R. 1987. Chronology REFERENCES of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. 12 1Ahmad, R., Babb, S., DeFour, J. & Shurland, D. 1986. Hudson, D. & Mountjoy, E.W. 1990. Patch reef ecology, The development of successive structures in the sedimentology and distribution Buccoo Reef complex, Mount Dillon Formation of Tobago: in Rodrigues, K. Tobago: in Larue, D.K. & Draper, G. (eds), Transac- (ed.), Transactions of the First Geological tions of the 12th Caribbean Geological Conference, St. Croix, U.S.V.I., August 7th-llth, 1989, 376-388. Conference of the Geological Society of Trinidad and 13 Tobago, Port-of-Spain,July 10th-12th, 1985, 40-52. Hudson, D.I.G. & Mukherji, K.K. 1985. Preliminary pe- 2Bachew, S. & Lewis, N. 1986. The impact of beach sand trology of Bon Accord beach sand, southwestern To- mining at Goldsborough Bay, Tobago: in Rodrigues, bago, West Indies. Transactions of the 4th Latin K. (ed.), Transactions of the First Geological Confer- American Geological Conference, Port-of -Spain, Trinidad, July 7th-15th, 1979, 179-196. ence of the Geological Society ofTrinidad and 14 Tobago, Port-of-Spain, July 10th-l2th, 1985, 78-84. Hudson, D.I.G., Mukherji, K.K., Sawh, H., Jenkins, J.T. 3Beckmann, J.P., Butterlin, J., Cederstrom, D.J., & Shevchenko, G. 1982. Beach sediments of Tobago, Christman, R.A., Chubb, L.J., Hoffstetter, R., Kugler, West Indies: in Snow, W., Gil, N., Llinas, R., Ro- H.G., Mar-tin-Kaye, P.H.A., Maxwell, J.C., Mitchell, driguez-Torres, R., Seaward, M. & Tavares, I. (eds), R.C., Ramirez, R, Versey, H.R., Weaver, J.D., Transactions of the 9th Caribbean Geological Confer- Westermann, J.H. & Zans, V.A. 1956. Lexique ence, Santo Domingo, Dominican Republic, August 16th-20th, 1980, 1, 139-149. Stratigraphique International, Volume 5. Amerique 15 Latine, 2b, Antilles. CNRS, Paris, 495 pp. Jackson, T.A., Duke, M.J.M., Smith, T.E. & Huang, C.H. 4Briden, J.C., Rex, D.C., Fuller, A.M. & Tomblin, J.F. 1988. The geochemistry of the metavolcanics in the 1979. K-Ar geochronology and palaeomagnetism of Parlatuvier Formation, Tobago: evidence of an island volcanic rocks in the Lesser Antilles island arc. Philo- arc origin: in Barker, L. (ed.), Transactions of the llth sophical Transactions of the Royal Society, London, Caribbean Geological Conference, Dover Beach, Bar- bados, July 20th- 26th, 1986,21.1-21.8. A291, 485-528. 16 5Cunningham-Craig, E.H. 1907. Preliminary Report by Jackson, T.A. & Smith, T.E. 1986. Metasomatism in the the Government Geologist on the Island of Tobago. Tobago Volcanic Group, Tobago, W.I.: in Rodrigues, Trinidad Council Paper, no. 9. [Not seen]. K. (ed.), Transactions of the First Geological Confer- 6Donnelly, T.W., Beets, D., Carr, M.J., Jackson, T., ence of the Geological Society of Trinidad and Tobago, Port-of-Spain, July 10th-12th, 1985, 34-39. Klaver, G., Lewis, J., Maury, R., Schellenkens, H., 17 Smith, A.L., Wadge, G. & Westercamp, D. 1990. Johnson, M.E. 1988. Why are ancient rocky shorelines so uncommon? Journal of Geology, 96, 469-480. History and tectonic setting of Caribbean magmatism: 18 in Dengo, G. & Case, J.E. (eds), The Geology of Leterrier, J., Maury, R.C., Thonon, P., Girard, D. & Mar- North America, Volume H, The Caribbean Region, chal, M. 1982. Clinopyroxene composition as a 339-374. Geological Society of America, Boulder, method of identification of the magmatic affinities of Colorado. paleo-volcanic series. Earth and Planetary Science 7 Letters, 59, 139-154. Donnelly, T.W. & Rogers, J.J. 1978. The distribution of 19 igneous rock suites throughout the Caribbean. Geolo- Lewis, D.N. & Donovan, S.K. 1991. The Pliocene Echi- gie en Mijnbouw, 57, 151-162. noideaof Tobago, West Indies. Tertiary Research, 12, 8 139-146. Donovan, S.K. 1989. Palaeoecology and significance of 20 barnacles in the mid-Pliocene Balanus Bed of Tobago, Maxwell, J.C. 1948. Geology of Tobago, British West West Indies. Geological Journal, 24, 239-250. Indies. Bulletin of the Geological Society of America, 9 59, 801-854. Frost, C.D. & Snoke, A.W. 1989. Tobago, West Indies, a 21 fragment of a Mesozoic oceanic island arc: petro- Morgan, F.D., Wadge, G., Latchman, J., Aspinall, W.P., chemical evidence. Journal of the Geological Society, Hudson, D. & Samstag, F. 1988. The earthquake haz-

206 T. A. JACKSON and S. K. DONOVAN

ard alert of September 1982 in southern Tobago. 30Snoke, A.W., Rowe, D.W, Yule, J.D. & Wadge, G. Bulletin of the Seismological Society of America, 78, 1991. Tobago, West Indies: a cross-section across a 1550-1562. fragment of the accreted, Mesozoic oceanic -arc, of 220'Brien, C.B. & Lawson, D. 1988. Nearshore processes the southern Caribbean: in Gillezeau, K.A. (ed.), and sedimentation at Queen's & Richmond Bays, To- Transactions of the Second Geological Conference bago, West Indies: in Barker, L. (ed.), Transactions of of the Geological Society of Trinidad and Tobago, the llth Caribbean Geological Conference, Dover Port-of -Spain, Trinidad, April 3-8, 1990, 236- 243. 31 Beach, Barbados, July 20th-26th, 1986,14.1-14.25. Snoke, A.W, Yule, J.D, Rowe, D.W. & Wadge, G. 1990. 23Rowe, D.W. 1987. Structural and petrologic history of Geologic history of Tobago, West Indies. Abstracts, northeastern Tobago, West Indies: a partial cross-sec- 2nd Geological Conference of the Geological Society tion through a composite oceanic arc complex. Unpub- of Trinidad and Tobago, Port-of-Spain, April 2nd-8th, lished M.Sc. thesis, University of Wyoming, Laramie. 19??, 17-19. 24Rowley, K. 1979. Field trips C and F Tobago. Field Guide, 32Snoke, A.W, Yule, J.D, Rowe, D.W, Wadge, G. & 4th Latin American Geological Conference, Port-of- Sharp, W.D. 1990. Stratigraphic and structural Spain, Trinidad, July 7th-15th, 65-69. rela tionships on Tobago, West Indies, and some 25Rowley, K. & Roobol, MJ. 1978. Geochemistry and tectonic implications: in Larue, D.K. & Draper, G. age of the Tobago igneous rocks. Geologie en (eds), Transactions of the 12th Caribbean Mijnbouw, 57, 315-318. Geological Conference, St. Croix, U.S.V.L,August 26Saunders, J.B. & Muller-Merz, E. 1985. The age of the 7th-11th, 1989,389-402. 35 Rockly Bay Formation, Tobago. Transactions of the Stanley, S.M. & Campbell, L.D. 1981. Neogene 4th Latin American Geological Conference, Port-of- mass extinction of western Atlantic molluscs. Spain, Trinidad, July 7th-15th, 1979, 1, 339-344. Nature, 293, 457-459. 34 27Sharp, W.D. & Snoke, A.W. 1988. Tobago, West Indies; Trechmann, CT. 1934. Tertiary and Quaternary beds geochronological study of a fragment of a composite of Tobago, West Indies. Geological Magazine, 71, Mesozoic island arc. Geological Society of America 421-493. 35 Abstracts with Programs, 20, 60. Wadge, G. & Hudson, D. 1986. Neotectonics of southern 28Snoke, A.W. & Rowe, D.W. 1986. Petrotectonic Tobago: in Rodrigues, K. (ed.), Transactions of the 1st setting of Tobago, West Indies. Geological Society of Geological Conference of the Geological Society of America Abstracts with Programs, 18, 756. Trinidad and Tobago, Port-of-Spain, July 10th-12th, 29Snoke, A.W., Rowe, D.W, Yule, J.D. & Wadge, G. 1990. 1985, 7-20. 36 New geological map of Tobago, West Indies: Yule, J.D, Snoke, A.W. & Rowe, D.W. 1988. A dy- summary and some important revisions. Abstracts, namothermal inverted metamorphic aureole on 2nd Geological Conference of the Geological Tobago, West Indies: a product of megathrusting Society of Trinidad and Tobago, Port-of -Spain, April or forceful intrusion? Geological Society of 2nd-8th, 17. America Abstracts with Programs, 20, 60-61.

207

208

CHAPTER 12

Trinidad

STEPHEN K. DONOVAN

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION STRATIGRAPHY: NORTHERN RANGE

TRINIDAD OCCUPIES a geologically important position The Northern Range of Trinidad is a subterrane of the at the southeast corner of the Caribbean Plate and north- Cordillera de la Costa Province of northeastern South Amer- eastern South America, including a fragment of the Meso- ica, produced by the emplacement south or southeast of zoic metamorphic belt (Fig. 12.1). The metamorphosed extensive sheets of Mesozoic metavolcanic s (only the Sans Mesozoic succession of the east-west trending Northern Souci Formation in Trinidad) and metasediments into Pa- Range of Trinidad forms part of the Araya-Paria-Trinidad leogene flysch21. The metamorphic grade is lower green- Province21, being geologically continuous with related schist facies2, with a generally east-west trend of foliations. rocks in northern Venezuela (see Donovan, this volume). The Northern Range is bordered to the north by the North The Northern Range is divided from the less deformed Coast Fault Zone and to the south by the El Pilar Fault Zone. Cretaceous and Cenozoic successions of the Trinidad Prov- The Mesozoic rocks of the Northern Range comprise the ince21 to the south by the major El Pilar Fault Zone. On- Caribbean Group and were first described by Wall and shore, the Trinidad Province is divided into four Sawkins82. physiographic regions, each with an approximately east- The stratigraphy of the Caribbean Group is imperfectly west trend (Fig. 12.2); the Northern Basin, the Central understood due to a combination of factors, particularly the Range, the Southern Basin and the Southern Range. poor exposure afforded by the dense forest cover, an incom- Including the Northern Range, these five physiographic plete understanding of the structural geology and a paucity regions each show individual rock successions and struc- of biostratigraphically useful fossils. Two contrasting li- tural styles. The stratigraphy of each of these regions is thostratigraphic schemes exist. The succession illustrated in considered in turn herein, traversing from north to south, Figure 12.3A is based on the supposition that the sequence followed by a discussion of the structural geology and a essentially youngs from north to south, with the Maracas synthesis of the geological history. There are a number of Formation being the oldest and younger beds in the north other published reviews of the geology of Trinidad, amongst being thrust onto older beds. However, Potter53 demon- which are recommended those by Kugler37, Suter 74, Barr strated that a major overturned anticlinal structure in the and Saunders11, and Carr-Br own and Frampton19. western Northern Range has led to a repetition of sequences. The sedimentary succession of Trinidad contains many Thus, the Maracas and Grande Riviere Formations (Fig. monotonous sequences of fine-grained clastic rocks that 12.3A) have been reinterpreted as the northern and southern have not been easily subdivisible on lithologic grounds. limbs, respectively, of the same formation within this anti- Biostratigraphy, particularly that based on foraminifers, is cline. Saunders and co-workers are preparing a revised map therefore of extreme importance in Trinidadian geology67. of the Northern Range, part of which (Fig. 12.3B) is in- However, detailed discussion of foraminiferal biostrati- cluded herein for comparison. That some confusion exists graphic schemes is outside the remit of the present discus- concerning the succession of strata is demonstrated by Frey sion and the interested reader is directed to the many relevant et al.24, who included the Rio Seco Formation on a map publications detailed in the reference list. while excluding it from their stratigraphic chart. The strati-

209

Trinidad

Figure 12.1. Regional geotectonic map of the southern Caribbean region showing the inferred distribution of fragments of the Cretaceous oceanic arc and associated geological terranes (reproduced by permission of the Geological Society from * Tobago, West Indies, a fragment of a Mesozoic oceanic island arc: petrochemical evidence" by Carol D. Frost and Arthur W. Snoke in Journal of the Geological Society, London, volume 146 for 1989). graphy discussed below corresponds to that in Table 12.1. able foraminiferans, which Furrer26 and Saunders64 considered Further discussion of those formations mentioned in Figure indicative of a shallow water depositional environment. 12.3A, but not discussed herein, can be found in references given in the text. Maracas Formation The Maracas Formation conformably overlies the Ma- Maraval Formation raval Formation. Potter53 considered that "The originally The Maraval Formation is the oldest formation within described outcrop is the overturned northern limb of the the western Northern Range. The base is not seen and the Northern Range anticline; the upright southern limb was minimum thickness of the unit is postulated to be 500 m. shown as the Grande Rivifere Formation by Kugler40." The Potter53 considered the formation to comprise three hori- Maracas Formation comprises interbedded quartzites and zons of massive, recrystallized limestone, including the phyllites with rare massive quartzites and slates, with a few development of a possible reef knoll, interbedded with thin limestones near the base and top. Sedimentary struc - calc -schists. Lenses of gypsum37 are also present. tures produced by turbiditic deposition52 indicate a palaeocurrent direction from north to south10,53. Volcanic ash Rio Seco Formation bands are present and a metavolcanic horizon of tholeiitic It is uncertain if the Maraval Formation should include affinity from near the base of the formation is chemically the Rio Seco Formation of the eastern Northern Range, or similar to the volcanics of the Sans Souci Formation32. whether both formations should be considered as separate units. A Tithonian (uppermost Jurassic) age has been deter- Chancellor Formation mined for the Rio Seco Formation on the basis of the Potter 53 considered the Chancellor Formation to rest occurrence of the ammonite Perisphinctes transitorius Op- conformably on the Maracas Formation, but to be unconfor- pel70. Potter52 considered the facies of this unit to include mably overlain by the Morvant Beds. The sedimentary massive limestones in the central Northern Range, flanked sequence of the Chancellor Formation comprises lower by calcareous phyllites. The dominant bioclasts are echinoid limestones (100 m), lower phyllites (40-180 m), upper lime- plates, with algae, rare molluscan fragments and question- stones (170 m) and upper phyllites (20-100 m)53. The re-

210

STEPHEN K. DONOVAN

crystallized limestones are less pure than those of the Mara- sedimentary basins containing Cretaceous carbonates and cas Formation. elastics, Paleogene flysch and neritic deposits, and Neogene marine and deltaic sediments. This sequence may total 10 Toco Formation km in thickness and is disrupted by mud diapirs. The Toco Formation of the eastern Northern Range The Cenozoic stratigraphic succession in the Trinidad comprises dark calcareous shales with thin limestones and Province is summarized in Figure 12.4. The Cretaceous occasional coarse-grained quartzites9. Trechmann76 de- succession of the same area, which only outcrops in the scribed a sponge-bearing reef limestone in Toco Bay which Central Range, is summarized in Table 12.2. Details of type provided the first evidence for a Cretaceous age for the localities are given in Kugler38. Although major strati- Northern Range9. The Tompire Formation, dated as Barre- graphic units are considered within separate physiographic mian in age on the basis of apyritized ammonite fauna31 and regions herein, there are at least time stratigraphic equiva- its included foraminiferans12,64, comprises calcareous, lents in adjacent areas until the Pliocene (Fig. 12.4). semi-phyllitic shales interbedded with occasional thin limestones in the type area at Tompire Bay9. However, the Toco Northern Basin and Tompire Formations can only be differentiated in This is a younger Cenozoic synclinal basin which un- coastal sections and are generally grouped together as a derlies the Caroni Plains, a low -lying area of terraces, allu- single unit. vial plains and swamps 11. The contact with the Northern Range is the El Pilar Fault Zone (Fig. 12.3A). Sans Souci Formation Cunapo Formation Apart from minor volcanic horizons in some other The Cunapo Formation is a sequence of conglomerates formations (see Maracas Formation, above), the Sans Souci that range from the Upper Oligocene to the Upper Pliocene, Formation represents the only outcrop of igneous rocks in interdigitating with various formations in the Northern Ba- Trinidad. The formation only outcrops between the villages sin19 (Fig. 12.4 herein). The Cunapo Formation is developed of Sans Souci and Grande Riviere on the eastern north coast. adjacent to the Northern Range, which acted as the source The formation comprises massive basaltic volcaniclastics, area during its uplift. The Cunapo, Brasso and younger basaltic lavas, intrusive gabbros and terrigenous sedimen- formations are Northern Basin units which onlap the Central tary rocks, including limestones, shales, sandstones and Range (R.D. Liska, written communication). 81 conglomerates . The contact with the Toco Formation to Brasso Formation the south is faulted. The inferred age of the Sans Souci 81 The Brasso Formation (Lower to Middle Miocene) has Formation is ?Aptian-?Santonian . a maximum thickness of about 1,800 m37 and is sub-divided into three principal members38: Galera Formation The Galera Formation comprises non-calcareous, dark - Navarro River Member (youngest) grey shales interbedded with occasional siltstones and 52 quartzitic sandstones . Sandstones predominate in the - Tunnel Hill Member lower part of the formation, the upper part of the section - Esmerelda Member (oldest) being shaley. Included foraminiferans are Upper Creta- ceous8,52 . The Esmerelda Member comprises dark bluish-grey to bluish-black silts and includes a rich fauna, including occa- Morvant Beds sional layers of well-preserved neritic pteropod molluscs. The Morvant Beds53 comprise black, pyritic, gypsifer- The Ste. Croix Member is a lateral equivalent of the Es- ous shales interbedded with coarse-grained calcareous sand- merelda Member, comprising foraminiferal silts and clays stones with wildflysch blocks and boulders, probably with minor sand beds, and yielding a fauna of corals, derived from the Chancellor Formation and Morvant Beds. pteropods and foraminiferans. The middle and upper parts The contact with underlying Chancellor Formation is un- of the Brasso Formation comprise grey clays and silts which conformable. include interbedded sands that may grade into conglomer- ates74. Higgins and Saunders29 noted that the different members of this formation are difficult to distinguish in the STRATIGRAPHY: TRINIDAD PROVINCE field without palaeontological evidence. The lower boundary of the Brasso Formation is not Case et al.21 summarized the geology of the Trinidad Prov- accurately known in the Northern Basin and may be under- 19 ince as comprising a series of superimposed and deformed lain by the Lower Cipero Formation , although Tyson and Ali78 have suggested that it is underlain by a sequence

211

Trinidad

Figure 12.2. Outline map of Trinidad showing the five principal physiographic regions (redrawn after Rodrigues60). comprising the Lower Cipero, Cunapo and Nariva Forma- Eocene age37. The fauna is dominated by the benthic tions. In the type section of the Brasso Formation, in the foraminifer Amphistegina and the alga Lithothamnia, with Central Range, the Ste. Croix Member rests on the Nariva some corals (such as Porites) and molluscs, although macro- Formation39 and different members are known to overstep fossils are poorly preserved. Lower limestones are rubbly, older formations down to the Paleocene38. Suter74 considered becoming more massive towards the top74. the lower boundary to be unconformable. The Brasso The limestones of the Tamana Formation have tradi- Formation interdigitates with the Cunapo Formation to the tionally been considered to be reefal in origin. Wharton et north and the Nariva Formation to the south. al.83 have shown the Brigand Hill Limestone of the north- The fauna of the Brasso Formation indicates an outer eastern Central Range to be a large algal shoal reef or a reef neritic to bathyal environment of deposition19,38. shoal. Tamana Formation Manzanilla Formation The Tamana Formation comprises massive, white to The Manzanilla Formation (Upper Miocene) is sub-di- yellow, granular to crystalline limestones with intercalated vided into three members19,29: sands and clays38. Various calcareous units are included within this formation, including the Guaracara (youngest; - Telemaque Sandstone Member (youngest) Middle Miocene), Biche and Brigand Hill (oldest; Lower - Montserrat Glauconitic ‘Sandstone’ Member Miocene) Limestones, the limestones becoming younger - San Jose Calcareous Silt Member (oldest) from east to west (R.D. Liska, written communication). Total thickness is approximately 100 m37,38. These limestones have sharp, discordant contacts with the underlying The San Jos6 Member comprises dark grey to black, clays and silts of the Brasso Formation74, or with the Nariva calcareous silts which have been burrowed. A rich fauna of Formation or sediments of Lower Cretaceous to Middle foraminiferans and small molluscs indicates an open mar rine, inner sublittoral environment of deposition63. The

212

STEPHEN K. DONOVAN

Montserrat Member is only locally developed It is not a true iron-cemented sands and conglomerates38. Clasts in the sandstone, but a glauconitic pelleted clay mixed with abun- conglomerates include chert, quartz and 'porcellanite'. dant shell debris . The sands of the Telemaque Member Some clays include plant debris. include arenaceous foraminiferans indicative of brackish conditions of deposition19. Shelly lenses within this mem- Central Range ber have produced a diverse mollusc fauna. The Central Range is a line of northeast-southwest The succession within the Manzanilla Formation indi- trending hills formed by an asymmetrical anticlinal structure cates infilling and shallowing of the Northern Basin. The with a Cretaceous (Table 12.2) to Paleogene (Fig. 12.4) diverse mollusc fauna differs from those of the underlying core11. 38 Brasso and the overlying Springvale Formations . The Cuche Formation Manzanilla Formation is the main producing horizon of the The oldest rocks outcropping in the core of the Central 19 North Soldado Field in the Gulf of Paria . This is the only Range form the Cuche Formation, although Tyson and Ali78 example of significant petroleum production from Northern 17 56 have shown it to be underlain by Couva Evaporites and Basin sediments . basal sands of equivalent age to the Maracas Formation of Springvale Formation the Northern Range. The Cuche Formation is divided into The Springvale Formation (Pliocene) has a total thick- two members38: ness of about 1,100 m and is divided into three mem- 19,29 bers . - Maridale Marl Member (younger) - La Carriere Shale Member (older) - Chickland Clay Member (youngest) - Savaneta Glauconitic Sandstone Member The La Carriere Member comprises dull grey, calcare- - Gransaull Clay Member (oldest) ous, silty shales with occasional thin sandstones and nodular ironstone bands, carbonaceous partings and anhydrite38,39. The Gransaull Member also includes common silts and The upper part of the member is an increasingly calcareous sands. The Savaneta Member is a thin, shelly sandstone that mudstone, with intercalated arkosic quartzites. Belemnites, 29 is a good marker bed in the western part of the island . rare fragments of ammonite and a good fauna of foraminif- The Springvale Formation is unconformable on the erans all indicate a Barremian age12,37. Rare, intraforma- Manzanilla Formation and is unconformably overlapped by tional conglomerates comprise massive limestone blocks 38 the Talparo Formation . The Springvale Formation se- occurring as slip masses which were derived from an quence represents a marine incursion after the shallowing Urgonian facies1,39, containing corals and caprinid 19 indicated by the Manzanilla Formation , although occa- rudists78. sional lignite horizons indicate the close proximity of The Maridale Member includes belemnite-bearing, 37 coastal swamps . The included faunas indicate an inner foraminiferal marls and calcareous mudstones37,38 and out- neritic environment. Mollusc-rich beds are locally devel- crops in the core of the Central Range. Contacts are always oped. faulted74. The foraminiferans indicate this member to be Talparo Formation Upper Aptian to Lower Albian13. The Cuche Formation is The Talparo Formation (Pliocene) is a sequence of over overlain unconformably by Upper Cretaceous shales and the 1,000 m of sands, sandy clays and clays with occasional Paleocene Chaudiere Formation38. lignites which were deposited under mar ginal marine to Gautier Formation 37,38 brackish and freshwater conditions . The formation be- Unconformably overlying the Cuche Formation, the 74 comes conglomeratic towards the Northern Range . The Gautier Formation represents an abrupt change of facies to formation is sub-divided into six members. Marine clays mainly well-bedded, black, bituminous and calcareous predominate in the lower part of the sequence. Spontaneous shales 39, with sandy conglomerates containing pebbles and ignition of lignites in the overlying sands and silts of brack- bivalves, and laminated sandstones and mudstones78. This ish water origin has baked adjacent fine-grained sediments 37 represents a deeper water facies and contains an abundant to bright, brick-red 'porcellanites' . The formation rests fauna of planktonic foraminiferans of Cenomanian age. The unconformably on the Springvale Formation and is overlain Gautier Formation is separated from the overlying Na- by Pleistocene river terraces. parima Hill Formation by a disconformity38. Cedros Formation Naparima Hill Formation The Pleistocene Cedros Formation comprises poorly- The Upper Cretaceous Naparima Hill Formation is a consolidated, blocky clays and fine- to coarse-grained sequence of well-bedded, sometimes bituminous mudstones sands, yellow, red and brown in colour, with lenses of and shales, with some marls and bituminous lime-

213

Trinidad

Table 12.1. Mesozoic formations of the Northern Range of Trinidad, based on Frey et al.24, Ahmad3, Wadge and Macdonald81, and Potter53,54. The given thicknesses are probably inaccurate due to structural repetitions2.

Formation Thickness Stratigraphic Position Lithology

Morvant Beds >220m ?Upper Cretaceous phyllites, sandstones, conglomerates

Galera <200m Campanian-Maastrichtian grits, conglomerates, sandstones ±non-calcareous shales

Sans Souci 1000 m ?Aptian-?Santonian basaltic flows and tuffs with a gabbroic sill, terrigenous sediments

Toco 1200m Barremian-Aptian black shales, sandstones, grits, conglomerates, limestone lenses

Chancellor 400-500 m Lower Cretaceous limestones, phyllites

Maracas 1200-1500 m uppermost Jurassic -lowermost non-calcareous phyllites, quartzites, Cretaceous limestones, calcareous phyllites, altered volcanic ash band

Rio Seco >450 m Tithonian limestones, calc -phyllites

Maraval >500 m Upper Jurassic limestones, calc -phyllites

37,38,78 stones .The upper part of this formation is a "...curi- Formation oversteps various Cretaceous formations. A ba- ous silicified siltstone/claystone referred to colloquially as sal conglomerate, the St. Joseph Boulder Bed, is developed, "argiline" or "argillite"..."11. Cherts are abundant in these 37 74 which Kugler considered as evidence of end Cretaceous upper layers . Biostratigraphic markers include both ben- orogenic activity. The Chaudiere Formation conformably thomc and planktonic foraminiferans. A deep water, low 19 merges upwards into the Pointe-a-Pierre Formation. energy environment of deposition is suggested . Pointe-a-Pierre Formation Guayaguayare Formation The Pointe-a-Pierre Formation comprises a lower The Maastrichtian Guayaguayare Formation is a dark shale-rich and an upper, coarse-grained sandy member38. grey, calcareous shale, disconformable on the Naparima Hill The sandstones show a range of sedimentary structures that Formation and unconformably overlain by the Lizard 38,39 indicate a turbiditic origin and individual beds can be up Springs Formation . Slip masses of Guayaguayare Formation 74 to 15 m thick . The suggested environment of deposition occur in the Paleogene Chaudiere Formation. is deeper water. Macrofossils are lacking, the characteristic Chaudiere Formation fauna comprising arenaceous foraminfers. This formation is The Chaudiere Formation comprises non-calcareous, 38 conformable on the Chaudiere Formation . dark green, olive green or grey, slightly silty shales which Navet Formation alternate with occasional coarse-grained sandstone layers, 37,38 The deep water, open marine facies of the Lizard particularly in the upper part . This unit is a ‘calcareous-type' flysch which includes large slip masses37. The Springs Formation (see below) continues into the Navet formation lacks macrofossils, but is typified by a rich fauna of Formation, which comprises up to about 300 m of light grey arenaceous foraminiferans19. The best development of this to greenish grey, sometimes chalky marls and calcareous sequence is near and over the Central Range uplift, where it clays with an abundant fauna of foraminiferans, particularly reaches a maximum thickness of circa 800 m7,39. The planktics. with occasional horizons rich in radiolari- ans19,37,38. The Navet Formation is subdivided into a number Chaudiere Formation was originally mapped in the Central 38 Range and has yet to be proven in drill cores in the Northern of marls which are not given member status . Basin (R.D. Liska, written communication). Nariva Formation The lower disconformable boundary of the Chaudiere The lower part of the Nariva Formation is a monotonous sequence of red-weathering mudstones and shales with

214

STEPHEN K. DONOVAN

Table 12.2. Cretaceous strata of the Trinidad Province, exposed as a series of inliers in the Central Range (Fig. 12.3A). Based on data in Kugler38, Higgins and Saunders29, Carr-Brown and Frampton19, Tyson and Ali78 and R.D. Liska (written communication). ______Formation Thickness Stratigraphic Position Lithology ______

Guayaguayare 120m Maastrichtian calcareous shale Naparima Hilt up to 600 m Upper Turonian- Campanian bituminous mudstones and shales, marls, silicified siltstones and mudstones, cherts Gautier up to 600m Upper Albian-Lower Cenomanian bituminous and calcareous shales, sandy conglomerates, mudstones Cuche 600-1500m Barremian- Albian calcareous shales, marls, sandstones, nodular ironstones, anhydrite, massive limestone slip masses, calcareous mudstones ______

occasional sandstone and mudstone lenses. The upper part The Southern Range is a series of low hills at the of the formation is mainly silts and sands, with thin lig- southern margin of the island, produced by a series of small, nites37. However, the upper part of the formation shows an discontinuous anticlinal structures which are mainly associ- unusual development of 'wildflysch', containing gravity ated with mud diapirism11. slide masses up to several hundreds of metres in length from While the petroleum geology of Trinidad is not specifi- all older Central Range formations37,39. Olistostromic cally discussed herein due to limitations of space, principal blocks from younger formations occur lower in the Nariva fields, trap geometry and producing horizons are summa- Formation than those of older formations, suggesting a rized in Table 12.3. progressive erosion39. Lizard Springs Formation The fauna is characterized by benthic, mainly aren- The Lizard Springs Formation is a dark green to grey, aceous foraminiferans and planktics are rare19. The sands compact to nodular, poorly stratified marl and calcareous may be proximal turbidites78 and the sequence probably clay, associated with wildflysch conglomerates and large represents a relatively deep water, turbid environment. On slip masses11,66, with a total thickness of greater than 300 the basis of the typical Nariva Formation foraminiferal m38. The formation is unconformable on the underlying biofacies, Liska43 considered this unit to be diachronous Cretaceous Guayaguayare and Naparima Hill Formations , within the Lower Miocene. but conformably grades into the Navet Formation. The The Nariva Formation interfingers with both the Brasso Chaudiere and Lizard Springs Formations interfinger, the Formation to the north and the Cipero Formation to the latter representing a more deep water environment of depo- south11. The base of the Nariva Formation is unconformable sition19. The Lizard Springs Formation is stratigraphically and conglomeratic, and the formation is in turn unconfor- subdivided on the basis of a rich fauna of planktic and mably overlapped by the Lengua Formation38. The sands of benthic foraminiferans. the Nariva Formation are stratigraphically the oldest major San Fernando Formation oil reservoirs in Trinidad, and are producing both on land The contact between the San Fernando and Navet For- and offshore in the Brighton area in the southwest part of mations is at the Mount Moriah Member, comprising bed- 19 the Central Range . ded silts and glauconictic sands overlain by a boulder bed29. The overlying beds of the San Fernando Formation include Southern Basin and Southern Range impure sandstones, silts, glauconitic shales and calcareous The Southern Basin is a synclinal structure with a foraminiferal clays, the last associated with the Vistabella Limestone, a lenticular biohermal unit comprising Li- fill of upper Cenozoic sediments. The complicated 37-39,74 folding and faulting of the rim of this basin has thothamnia algae and orbitoidal foraminiferans . produced many of Trinidad's oil-producing structures 11. Olistostromic blocks are also present. The formation is at The physiographic expression of the Southern Basin is a least 90m thick. low-lying, undulating countryside with some swamps The lithological contrast with the underlying Navet and, in the west at Brighton, the La Brea pitch lake. Formation is interpreted as being due to renewed tectonism

215

216

STEPHEN K. DONOVAN

producing rapid uplift in the basin. Thus, the highest marl of the Navet Formation, the Hospital Hill Marl, deposited in an open marine environment, is overlain by neritic nearshore facies of the basal Mount Moriah Member, suggesting that tectonic uplift was a geologically rapid event19,37. Cipero Formation The Cipero Formation lies unconformably upon the San Fernando Formation. The Cipero Formation comprises deep-water calcareous clays and marls with rich, dominantly planktic foraminferal faunas37. Estimates for maximum formation thickness vary up to 2,700 m37. A major included slip mass, the Bamboo Silt, is derived from the San Fernando Formation38. The Cipero Formation represents a return to deeper water, open sea conditions of deposition following a hiatus in sedimentation following deposition of the San Fernando Formation19. The Ciper o Foimation is widespread throughout the Southern Basin11. The divisions between the Lower and Middle, and Middle and Upper Cipero Formation are determined biostratigraphically at the Oligocene-Miocene and the Lower-Middle Miocene boundaries, respectively (Fig. 12.4). The Retrench Sands of the Middle Cipero Formation may have originated as deep water fans and are associated with radiolarian marls19,74. The younger (Upper Cipero) Herrera Sands are more important as an oil producer. The Herrera Sands have a well-documented turbiditic origin, with a source to the north or northwest30,36,51. Lithologi- cally, the Herrera Sands have a 'pepper and salt’ appearance, comprising fragments of chert, black shale, black limestone and coal fragments with clear quartz grains 30. The sands occur as narrow, elongate turbidite fans, trending northeast- southwest36; for example, the "intermediate" sand of Hosein30 is 27 km long by 615 m wide. Hosein30 considered that the Herrera Sands occur at three levels ("underthrust", "intermediate", "overthrust") separated by thrust faults. The Herrera Sand facies persists into the lower part of the Karamat Formation, but becomes siltier33. The Rio Claro Boulder Bed, formerly considered to be the basal unit of the Lengua Formation, has been shown by 43 Liska to be tectonic, not sedimentary, in origin and is Figure 12.3. (A) (opposite): Simplified geologi- included on foraminiferal evidence within the Cipero For- cal map of Trinidad (redrawn, with minor modi- mation. fication, after Barr and Saunders11; see also Karamat Formation 37,40 19 The Karamat Formation is a non-calcareous, greenish- Kugler ; Carr-Brown and Frampton ). Ages of 38 various formations discussed in the text and in grey clay . Silty sand lenses in the lower part of the Figure 12.4. Key: B-B=Los Bajos Fault; E- formation are late Herrera Sand units (see above). E=E1 Pilar Fault Zone. (B) (above): Geological Included olistoliths are of Upper Cretaceous to Lower Miocene antiquity19. The distinctive fauna comprises map of the Northern Range north of Port-of- 19 Spain, redrawn after Frey et al.24 (see also Pot- arenaceous foraminiferans . The formation has a ter53). Stratigraphy as in Table 12.1. Inset map maximum thickness of about 1,200 m. The lack of shows the part of northern Trinidad in main calcium carbonate is in contrast to the underlying Cipero diagram. and overlying Lengua Formations.

217

Trinidad

Lengua/Lower Cruse Formation clays. The reservoirs of the Cruse Formation are important The Lengua and Lower Cruse Formations were oil producers. grouped together as one unit by Liska42 on biostratigraphic Forest Formation and lithologic grounds. The Upper Miocene Lengua Lower The Forest Formation comprises the Lower Forest Clay Cruse Formation is a sequence of dark greenish-grey, some- Member overlain by the Forest Sands Member. The Lower times gypsiferous calcareous clays and marls with a rich Forest Clay Member comprises pure to slightly silty clays 16 fauna of calcareous and arenaceous benthic and planktic with a fauna of arenaceous foraminiferans . This represents a 57 foraminfers (Lengua Formation38,39) overlain by greasy, regionally transgressive mud unit . The Lower Forest non-calcareous clays with claystone nodules and occasional Clay Member is unconformably overlain by the Forest sand layers, and including a fauna of arenaceous foraminif- Sands Member, which comprises massive to thinly bedded erans (Lower Cruse Formation). This unit rests conformably sands separated by silty clays and with occasional lignite 16 on the Karamat Formation. The Lengua Formation may be beds . This unit was deposited in a delta front/barrier bar 16,57 up to about 600 m thick38, overlain by up to 1,200 m of environment . The fauna includes both calcareous and Lower Lengua Formation clays39. Both marls and clays are arenaceous foraminiferans. The upper part of the Forest incompetent, and are associated with mud tectonism. Len- Sand Member is a major oil producing horizon. The Forest gua Formation foraminiferans indicate bathyal or deeper Formation is strongly unconformable with, and overlaps environments (R.D. Liska, written communication), while onto, the Cruse Formation74. the somewhat different fauna of the Lower Cruse Formation Morne l'Enfer Formation may indicate increased freshwater discharge prior to deltaic The Morne 1'Enfer Formation comprises four mem- deposition14. bers16: Roseau Formation The Roseau Formation is a grey, non-calcareous clay - Upper Morne l’Enfer Member (youngest) of deeper water origin with a similar lit hology and - Lot 7 Silt foraminifer fauna to the Lower Cruse Formation clay, but - Lower Morne 1'Enfer Member was formerly considered to be a lateral equivalent to the 14,19 42,44 - Upper Forest Clay Member (oldest) Cruse and Forest Formations . However, Liska pointed out that the Lower Cruse Formation is Late Miocene in age, whereas the Cruse and Forest Formations are both The Upper Forest Clay Member has a basal clay se- Pliocene, and considered the Roseau Formation to be an quence which grades up into a silty clay followed by alter- arenaceous facies of the Lengua/Lower Cruse Formation44. nating thin sands, silts and silty clays, with a good fauna of The fauna of the Roseau Formation includes agglutinated arenaceous and calcareous foraminiferans. This is overlain foraminiferans, with mangrove pollen as the dominant pa- by the Forest Silt at the base of the Lower Morne 1'Enfer lynomorph14. Member, comprising silts and silty clays with occasional Cruse Formation sands. These grade up into massive sands, well-bedded to In contrast to the clay facies of the underlying Len- cross-bedded, with thinner clay partings and including a gua/Lower Cruse Formation, the Cruse, Forest and Morne poor fauna of brackish water foraminiferans. The overlying 1'Enfer Formations have basal clays overlain by sandy se- Lot 7 Silt is a grey silt to silty clay, often with white laminae quences, a succession associated with the progressive shal- of fine-grained sandstone. At the top of the formation the lowing of the Southern Basin11. These formations represent massive sands of the Upper Morne 1'Enfer Member are three cycles of deltaic infilling of the Southern Basin, associated with grey silts and clays, lignitic clays, lignites with interbedded clays (up to 1.5 m thick) and porcellanites, younger formations being sandier and originating in more 16 shallow water19. with a poor fauna of brackish water foraminiferans . The The Cruse Formation is a sequence comprising about Morne 1'Enfer Formation rests unconformably on the Forest Formation. The total thickness of the formation may be 750 m or more of lenticular often channeled sands, inter- 15 bedded with clays and silts16,37. Sands are fine-grained and about 900 m. Bertrand considered the Morne 1'Enfer quartzitic, often massive and including a range of sedimen- Formation to consist of sheet sands deposited in a continental to inner fringe deltaic environment. A detailed palaeoen- tary structures such as basal mud clasts, small scale current 68 bedding, load casts, abundant submarine slumps and fine vironmental analysis by Saunders and Kennedy of the parallel laminations11,14. Some channels cut down into the upper part of this formation recognised a nearshore lagoonal Lower Cruse clays 37. Bertrand15 considered the Cruse For- environment that was infilled to tidal flat conditions and mation to be a paralic deposit with channel, barrier bar and succeeded by marsh or swamp. delta fringe sands, separated by transgressive and swampy

218

STEPHEN K. DONOVAN

Table 12.3. Oilfields of Trinidad (simplified after Persad47; see also Persad49 and Rohr61). The Moruga Group comprises the Gros Mome and Mayaro Formations (see text). The Catshill Formation is an equivalent of the Cruse and Forest Formations

Oilfield Name Producing Structure (Trap) Producing Formation

ONSHORE Balata-Bovallius detached overthrust Herrera Sands Barrackpore-Penal detached overthrust Herrara Sands Brighton-Point Ligoure detached overthrust/strike slip-en echelon Morne 1'Enfer anticline Catshill strike slip-en echelon anticline Catshill Coora-Quarry-Mome Diablo strike slip-stratigraphic strike Cruse-Forest-Morne 1'Enfer Fyzabad-Forest Reserve slip-en echelon anticline Cruse-Forest Guayaguayare Group detached growth fault related Moruga Moruga East strike slip-anticline Forest equivalent Oropouche detached overthrust Retrench Sands Palo Seco-Erin-Grande Ravine strike slip-stratigraphic Cruse-Forest-Morne 1'Enfer Point Fortin-Parrylands-Guapo strike slip-anticlines Cruse-Forest-Morne 1'Enfer Rock Dome East (Moruga North- detached overthrust Herrera Sands Trinity-Inniss) Rock Dome West (Moruga detached overthrust Herrera Sands West-Rock Dome)

OFFSHORE: GULF OF PARIA Soldado Main strike slip-anticline Cruse Soldado North strike slip-graben Manzanilla Soldado East strike slip Cruse-Forest-Morne 1'Enfer

OFFSHORE: EAST COAST Galeota Complex strike slip-anticline/detached overthrust Moruga and equivalent Poui detached growth fault/strike slip anticline Moruga and equivalent Samaan detached growth fault/strike slip anticline Moruga and equivalent Teak detached growth fault/strike slip anticline Moruga and equivalent

219

Trinidad

Gros Morne Formation (particularly strike-slip movements) and further disrupted The Gros Morne and Mayaio Formations form the by mud diapirism. The physiographic regions of the island 11 Moruga Group of southeastern Trinidad, which has a maxi- reflect this gross stmcture . A simplified map of the geo- mum thickness of between 1,500 m38 and 2,400 m74. This logical structure of Trinidad is given in Figure 12.5. group is predominantly composed of silty clays, silts and fine-grained sands14, foraminiferans, palynomorphs, plant Northern Range debris and rare molluscs are concentrated in the clays. The Northern Range is bounded by the North Coast In stratigraphic order, the Gros Morne Formation com- Fault Zone (NCFZ) and the El Pilar Fault Zone (EPFZ), both prises the Gros Morne Silts, the Gros Morne Sands, the St. major tectonic boundaries within the Southern Caribbean 4 Hilaire Silts and the Trinity Hill Sands19, deposited in a Plate Boundary Zone (Figs 12.1,12.5 herein). The contact nearshore marine environment23. This formation is an im- between the Northern Range and the younger Tertiary sedi- 11 portant oil producer in southeast Trinidad and off the east ments to the south is obscured by terraces and alluvium , coast. but is generally considered to be the EPFZ. However, 71 Mayaro Formation Speed has suggested that the Northern Range has been The Mayaro Formation is unconformable on the underthrust over the Caroni Syncline and is not bounded by the lying Gros Morne Formation. Telemaque75 considered the El Pilar Fault. 2,3 4,5 Mayaro Formation to include three members: Ahmad , and Algar and Pindell , have recognised three phases of deformation in the rocks of the Northern - Goudron Sand Member (youngest) Range. D1 structures were produced during the main phase - Mayaro Clay Member of deformation and metamorphism, producing tight folds with a north-directed asymmetry. These F1 folds are the key - Mayaro Sand Member (oldest) structural elements of the Northern Range and include the The Goudron Sand Member comprises a sequence of major anticlinal structure overturned to the north that was fine-grained, non-calcareous sands with associated peats 53,55 recognised by Potter . D structures overprint those of and lignites. The sands show a variety of shallow water 2 D , refolding F folds and with an associated S crenulation sedimentary structures and the included fauna indicates 1 1 2 cleavage, but are less pervasive. D structures are mainly brackish-water deposition. This member is over 900 m 3 broad and open F folds, which locally overprint D and D thick. The formation was deposited in a nearshore marine 3 1 2 structures, with associated faulting. environment. Palmiste Formation El Pilar Fault Zone The Palmiste Formation is a sequence of clays and The EPFZ and the NCFZ (Fig. 12.5) form part of the shales, with characteristic lignite beds and associated por- 20,38 San Sebastian-El Pilar right-slip transform system of eastern cellanites . The total thickness is over 900 m. The Palm- Venezuela and Trinidad21. Estimates for right-slip move- iste Formation unconformably overlies the Mayaro ment on this system vary between 0 and about 475 km69 Formation and is in turn unconformable beneath the overlying Quaternary deposits. Erin Formation The Erin Formation represents the final infilling of the Figure 12.4. (opposite) Onshore Cenozoic strati- western part of the Southern Basin19. This unit rests uncon- graphy of the Trinidad Province, based largely on 19 47 formably on the underlying Morne 1'Enfer Formation46, but Carr-Brown and Frampton and Persad (see also 78 is similar in sedimentary character11. It comprises thick, Tyson and Ali ), with important strati-graphic 44 bedded sands with interbedded grey silts and silty clays, revisions after Liska (written communication). lignitic clays, lignites up to 1 m thick, porcellanite, and Relative durations of geological epochs after 27 silicifed wood and iron pan near the base11,16. The thickness Harland et al. , although the length of the 46 reaches about 900 m in some areas . Pliocene is slightly exaggerated so that all formations within this interval could be included. Key: Pleisto=Pleistocene; W=west; C=central; 1 STRUCTURAL GEOLOGY SE=southeast; =includes northern flank of Cen- 2 19 tral Range; =Carr-Brown and Frampton sug- Suter 74 observed that the geological structure of Trinidad gested that the Brasso Formation was underlain comprises a series of thrust fault and fold belts arranged by Lower Cipero Formation, while Tyson and 78 parallel and en echelon, cut by a number of major faults Ali considered it to be underlain by the Cipero, Cunapo and Nariva Formations.

220

STEPHEN K. DONOVAN

221

Trinidad

Figure 12.5. Structural geology of Trinidad (redrawn after Ahmad3).

KEY C = upper Cenozoic sediments T = lower to middle Cenozoic sediments K = unmetamorphosed Cretaceous sediments M = metamorphosed Mesozoic sediments CRFZ = Central Range Fault Zone AF = Arima Fault EPFZ = El Pilar Fault Zone CF = Chupara Fault NCFZ = North Coast Fault Zone GRF = Grande Rivtere Fault SCFZ = South Coast Fault Zone

Throughout most of its length in Trinidad and Venezuela, the suggested that the plate boundary zone in the southeast EPFZ separates contrasting geological terranes80. In Caribbean may be 250-300 km wide, extending from Trinidad the EPFZ is considered to occupy a sheared and the Orinoco Delta in the south to Grenada in the north. downfaulted region occupied by synclinal and anticlinal Relative plate movements may thus be taken up on structures south of the ArimaFault3. Robertson et al59. have multiple fault systems and it would thus not be demonstrated that the EPFZ was an active right-lateral necessary for all faults to be simultaneously active. strike-slip system during the Pleistocene. However, geo- They have also postulated that strike-slip movements physical evidence does not support it being an active seismic along the NCFZ, the EPFZ, the Central Range Fault zone or strike-slip plate margin in Trinidad at the present Zone and the South Coast Fault Zone have juxtaposed day62, despite the evidence for it being an active strike-slip dissimilar stratigraphic sequences in adjacent dep- fault zone in Venezuela. Robertson and Burke58,59 have ositional basins.

222

STEPHEN K. DONOVAN

Trinidad Province Pleistocene65, producing associated shallow -rooted folds Barr and Saunders11 summarized the geological struc- which are major petroleum traps. Displacement on the fault ture of Trinidad south of the Northern Range as comprising is about 10 km85 . Tyson77 recognised both transtensional folds with a southwardly-directed asymmetry, that have and transpressional effects in regions where the fault di- been overfolded and overthrusted to the south. Middle Mio- verged from its general trend of 113° from north, producing cene tectonic movements separated the Northern and South- graben and anticlinal structures, respectively. ern Basins by the barrier of the Central Range79. To the south, the Southern Range Anticline is composed of a discontinuous line of anticlinal folds, thrust faulted and deformed by mud diapirs11,74. These anticlines Northern Basin strike out to sea at Erin Bay on the south coast of the The principal structural feature of the Northern Basin southwest peninsula and sometimes receive surface expres- is the Caroni Syncline (Fig. 12.5), which comprises gentle sion as ephemeral mud islands28. Mud diapir anticlines are folds apparently not affected by major thrust faults3. Sub- 11 elongate and narrow, ruptured at the crest of the uplift by sidiary anticline structures interrupt the southern flank , parallel strike faults or diapiric core faults74. The ejecta from with dips reaching up to 45° towards the Central Range. mud volcanoes and flows form a variety of geomorphic Central Range features. For general discussions of mud volcanism in southern Trinidad, see Higgins and Saunders28, Kugler41 and The structure of the Central Range was summarised by 34 74 Kerr et al. , amongst others. Suter as having a simple northern flank, a complexly The La Brea pitch lake is situated on the southwest faulted core and a southern flank which overthrusts to the 37 peninsula of Trinidad and represents an unusual type of south . In the east there is a sharp unconformity between diapirism. It had an original area, before exploitation, of the Cretaceous plus older Tertiary and the younger Tertiary, 6 11 about 0.5 sq km . The asphalt of the lake comprises 40% although the pre-Tertiary core is lacking in the west . bitumen, 30% mineral matter and 30% free water plus gas, Folding of these sediments may be related to movement on 39 the Central Range Fault Zone (Fig. 12.5), which shows both which Kugler interpreted as a mixture of Nariva Forma- tion sediments thickened by asphaltic oil. The lake is not strike-slip and thrust movements. The core of the Central 22 Range has been thrust southwards74. bowl-shaped, but irregular , varying in depth from 20 m in the centre to 1 km in the west. Two intersecting faults act as Immediately south of the Central Range proper is the Naparima Thrust and Fold Belt, a sequence of stronglv conduits for the asphalt. folded and faulted Tertiary sediments about 6-10 km wide11. Folds are clustered, narrow and elongate, often overturned GEOLOGICAL HISTORY towards the south and with limbs imbricately underthrust74. Structures are discontinuous, except for a series of south- 84 48 Wielchowsky et al. and Persad have determined a wardly overthrust anticlines. framework for the geological evolution of Trinidad, based on sedimentary and tectonic evidence, in which Southern Basin and Southern Range four phases of development are recognised. Structural styles south of the Central Range are defined by thrust faults (frequently associated with folds), clay Middle Jurassic to late Cretaceous: rift/passive margin diapir anticlines and associated mud volcanism, with an 3 phase overall southwardly-directed asymmetry . This phase was related to the formation of the proto- The Siparia-Ortoire Syncline is the major structural 11 Caribbean oceanic crust, following the Mesozoic separation component of the Southern Basin , and it is within the of the North and South American Plates18,50. Subsequent to folded and faulted sediments of this basin that Trinidad's this rifting episode, northern South America was a passive major oil fields are developed. The Fairylands Fold Zone, margin, with shelf sedimentation of shales and carbonates, to the north of the Los Bajos Fault in the southwest penin- a shelf-edge carbonate platform and turbidites in deeper sula, is an oil-producing region of shallow structures, par- water to the north48. A source to the south or southwest ticularly asymmetrical anticlines, developed in Neogene 74 provided sediment for basins in central and southern Trini- sediments . dad84. There is stratigraphic and structural continuity be- The Los Bajos Fault is a major vertical, right lateral tween rock sequences of this age in Venezuela and Trinidad. strike-slip fault which runs approximately eastsoutheast- These terrenes south of the EPFZ may be parautochthonous, westnorthwest across the southwest peninsula of Trinidad 3,11,77 whereas terranes north of the EPFZ are almost certainly and cuts across the Siparia-Ortoire Syncline . Move- allochthonous, having been transported tectonically. Speed ment on this fault occurred between the early Pliocene and

223

Trinidad

and Poland72 considered that these rocks of the Northern REFERENCES

Range are the deeper succession of a complex sedimentary 1 sequence, deposited in deep water marine environments in Ager, D.V. 1981. The nature of the stratigraphic record (2nd edition). Macmillan, London, xiv+122 pp. the Mesozoic and ?Paleogene as turbidites and hemipelagic 2 sediments on oceanic crust. Ahmad, R. 1988. Kinematic history of the Northern Range, Trinidad. Abstracts, Recent Advances in Caribbean Latest Cretaceous to mid-early Miocene: episodic Geology, Kingston, Jamaica, 18th-20th November, 1-2. 3 transpressional phase Ahmad, R. 1991. Structural styles in Trinidad: in Gillezeau, Deformation during this interval in Trinidad followed K. A. (ed.), Transactions of the Second Geological Con- two structural styles84: ference of the Geological Society of Trinidad and To- bago, Port-of- Spain, Trinidad, April 3-8, 1990, (A) strike-slip and oblique subduction of continental and/or 244-265. 4 oceanic lithosphere to the north, and probably involving Algar, S.T. & Pindell, J.L. 199la. Structural development of basement; and the Northern Range of Trinidad, and implications for the (B) 'thin-skinned' thrusting and folding to the south. tectonic evolution of the southeastern Caribbean: in Gillezeau, K. A. (ed.), Transactions of the Second Geo- Uplift and erosion to the north led to the deposition of logical Conference of the Geological Society of Trinidad clastic sequences in foreland troughs. and Tobago, Port-of-Spain, Trinidad, April 3-8, 1990, 6- 22. Mid-early Miocene to late middle Miocene: extensional 5Algar, S.T. & Pindell, J.L. 1991b. Stratigraphy and sedi- phase mentology of the Toco region of the Northern Range, N.E. Extension was probably related to large-scale transten- Trinidad: in Gillezeau, K. A. (ed.), Transactions of the sional forces as Trinidad and the Lesser Antilles Arc moved Second Geological Conference of the Geological Society past each other, resulting in the formation of the Northern of Trinidad and Tobago, Port-of-Spain, Trinidad, April 3- Basin. 8, 1990, 56-69. 6Barker, M.H.S. & Roberts, K.H. 1968. Excursion 6A south Late middle Miocene to present: transpressional phase Trinidad: in Saunders, J.B. (ed.), Transactions of the Local transtension emplaced and uplifted the regionally Fourth Caribbean Geological Conference, Port-of- metamorphosed rocks of the Northern Range, inverting the Spain, Trinidad, 28th March-12th April, 1965, 438- southern edge of the Caroni Syncline, and shortening un- 441. metamorphosed sedimentary sequences of Lower Creta- 7Barr, K.W. 1952. Limes tone blocks in the Lower Cretaceous ceous to Miocene rocks in central and southern Cuche Formation of the Central Range, Trinidad, B.W.I. Trinidad79,84. The newly-formed Northern and Southern Geological Magazine, 89, 417-425. Basins were separated by the Central Range, a region of 8Barr, K.W. 1966. A cross-section through the Northern erosion, with sediment being shed into the Southern Basin, Range of Trinidad: in Robinson, E. (ed.), Transactions of and subsequent biohermal growth. The end Miocene shal- the Third Caribbean Geological Conference, Kingston, lowing of the Southern Basin was due to infilling by deltaic Jamaica, 2nd-llthApril, 1962, 27-33. Geological Survey sediments derived from South America The sedimentary Department, Kingston. source for the infilling of the Northern and Southern Basins 9Barr, K.W. 1968. Excursion no. 2 eastern Northern Range and 45 73 may have been the proto-Orinoco or proto-Essequibo Toco district: in Saunders, J.B. (ed.), Transactions of the Rivers. Fourth Caribbean Geological Conference, Port-of-Spain, Trinidad, 28th March-12th April,1965, 430-432. ACKNOWLEDGEMENTS—I thank the Geological Society, London, and 10 Dr. A.W. Snoke for granting permission to reproduce Figure 12.1. Field- Barr, K.W. 1985. Graded bedding and associated phe- work in Trinidad was supported by grants from the Geologists' Association nomena in the Northern Range of Trinidad: in Trans- (G.W. Young Fund and Curry Fund) and the Geological Society of Trinidad actions of the Fourth Latin American Geological and Tobago, which I gratefully acknowledge. I thank Kirton Rodrigues and Conference, Port-of-Spain, Trinidad, 7th-15th July, particularly Bob Li ska for their valuable review comments on this paper. 1979,1,117-135. 11Barr, K.W. & Saunders, J.B. 1968. An outline of the geology of Trinidad: in Saunders, J.B. (ed.), Transac- tions of the Fourth Caribbean Geological Conference, Port-of-Spain, Trinidad, 28th March-12th April, 1965,

224

STEPHEN K. DONOVAN

1-10. 1990, 49. 12Bartenstein, H., Bettenstaedt, F. & Belli, H.M. 1957. Die 23Farfan, P. & Bally, K. 1991. An outline of the geology of Foraminiferan der Unterkreide von Trinidad, B.W.I. the Samaan Field, Trinidad West Indies: in Gillezeau, Erster Teil: Cuche- and Toco-Formation. Eclogae K.A. (ed.), Transactions of the Second Geological Con- Geologicae Helvetiae, 50, 5-67. ference of the Geological Society of Trinidad and To- 13Bartenstein, H., Bettenstaedt, F. & Bolli, H.M. 1966. Die bago, Port-of- Spain, Trinidad, April 3-8, 1990, Foraminiferen der Unterkreide von Trinidad, W.I. 139-156. Zweiter Teil: Maridale-Formation (Typlokalitat). 24Frey, M., Saunders, J. & Schwander, H. 1988. The min- Eclogae Geologicae Helvetiae, 59, 129-178. eralogy and metamorphic geology of low-grade me- 14Batjes, D.A.J. 1968. Palaeoecology of foraminiferal as- tasediments, Northern Range, Trinidad. Journal of the semblages in the late Miocene Cruse and Forest Forma- Geological Society, London, 145, 563-575. tions of Trinidad, Antilles: in Saunders, J.B. (ed.), 25Frost, C.D. & Snoke, A.W. 1989. Tobago, West Indies, a Transactions of the Fourth Caribbean Geological Con- fragment of a Mesozoic oceanic island arc: petrochemi- ference, Port-of-Spain, Trinidad, 28th March-12th cal evidence. Journal of the Geological Society, Lon- April, 1965, 141-156. don, 146, 953-964. 15Bertrand, W.G. 1985. Geology of the Point Fortin Field: 26Furrer, M.A. 1968. Paleontology of some limestones and in Transactions of the Fourth Latin American Geologi- calcphyllites of the Northern Range of Trinidad, West cal Conference, Port-of-Spain, Trinidad, 7th-15th July, Indies: in Saunders, J.B. (ed.), Transactions of the 1979,2, 690-699. Fourth Caribbean Geological Conference, Port-of- 16Bower, T.H. & Hunter, V.F. 1968. Geology of Texaco Spain, Trinidad, 28th March-12th April, 1965,21-24. Forest Reserve Field, Trinidad, W.I.: in Saunders, J.B. 27Harland, W.B., Armstrong, R.L., Cox, A.V., Craig, L.E., (ed.), Transactions of the Fourth Caribbean Geologi- Smith, A.G. & Smith, D.G. 1990. A geologic time scale cal Conference, Port -of-Spain, Trinidad, 28th March- 1989. Cambridge University Press, Cambridge, nth April, 1965,75-86. xvi+263 pp. 17Bray, R. & Eva, A. 1989. Age, depsoitional environment, 28Higgins, G.E. & Saunders, J.B. 1967. Report on the 1964 and tectonic significance of the Couva marine eva- Chatham Mud Island, Erin Bay, Trinidad, West Indies. porite, offshore Trinidad: in Duque-Caro, H. (ed.), American Association of Petroleum Geologists Bulle- Transactions of the Tenth Caribbean Geological Con- tin, 51,55-64. ference, Cartagena de Indias, Columbia, 14th-22nd 29Higgins, G.E. & Saunders, J.B. 1968. Excursion no. 4 August, 1983, p. 372. INGEOMINAS, Bogota. western Central Range: in Saunders, J.B. (ed.), Trans- 18Burke, K., Cooper, C., Dewey, J.F., Mann, P. & Pindell, actions of the Fourth Caribbean Geological Confer- J.L. 1984. Caribbean tectonics and relative plate mo- ence, Port-of-Spain, Trinidad, 28th March-12th April, tions. Geological Society of America Memoir, 162, 1965,435-437. 31-63. 30Hosein, F. 1990. Exploitation of the Middle Miocene oil 19Carr-Brown, B. & Frampton, J. 1979. An outline of the bearing intermediate Herrera Sandstone reservoirs in stratigraphy of Trinidad. Field Guide, Fourth Latin Trinidad, West Indies: in Larue, D.K. & Draper, G. American Geological Conference, Port-of-Spain, (eds), Transactions of the Twelth Caribbean Geologi- Trinidad, 7th-15th July, 12pp. cal Conference, St. Croix, U.S. Virgin Islands, 7th-llth 20Carr-Brown, B. & Young-On, V. 1990. Age and strati- August, 1989, 430-443. Miami Geological Society, graphic status of the Palmiste Formation, Trinidad, Florida. West Indies: in Larue, D.K. & Draper, G. (eds), Trans- 31Imlay, R.W. 1954. Barremian ammonites from Trinidad, actions of the Twelth Caribbean Geological Confer- British West Indies. Journal of Paleontology, 28,662- ence, St. Croix, U.S. Virgin Islands, 7th-llth August, 667. 1989, p. 534. Miami Geological Society, Florida. 32Jackson, T.A., Smith, T.E. & Duke, M.J.M. 1991. The 21Case, I.E., Holcombe, T.L. & Martin, R.G. 1984. Map of geochemistry of ametavolcanic horizon in the Maracas geologic provinces in the Caribbean region. Formation, Northern Range, Trinidad: evidence of Geological Society of America Memoir, 162, 1-30. ocean floor basalt activity: in Gillezeau, K.A. (ed.), 22Chaitan, W.B. & Graterol, V.R. 1991. A gravity investi- Transactions of the Second Geological Conference of gation of the pitch lake of Trinidad and Tobago: in the Geological Society of Trinidad and Tobago, Port- Gillezeau, K. A. (ed.), Transactions of the Second Geo- of -Spain, Trinidad, April 3-8, 1990, 42-41. logical Conference of the Geological Society of Trini- 33Jones, H.P. 1968. The geology of the Herrera Sands in the dad and Tobago, Port-of- Spain, Trinidad, April 3-8, Moruga West Oilfield of south Trinidad: in Saunders,

225

Trinidad

J.B. (ed.), Transactions of the Fourth Caribbean Geological Conference, Port-of-Spain, Trinidad, 7th- Geological Conference, Port-of-Spain, Trinidad, 15th July, 1979,2, 726-737. 28th Marcel2th April 1965, 91-100. 47Persad, K.M. 1985. Outline of the geology of the 34Kerr, P.P., Drew, I.M. & Richardson, D.S. 1970. Mud Trinidad area: in Transactions of the Fourth Latin volcano clay, Trinidad, West Indies. American Asso- American Geological Conference, Port-of -Spain, ciation of Petroleum Geologists Bulletin, 54, 2101- Trinidad, 7th-15th July, 7979,2,738-758. 2110. 48Persad, K.M. 1988. Tectonosedimentary evolution 35Kirmani, K.U. 1985. Geology of the Inniss, Catshill and of Trinidad and Tobago. Abstracts, Recent BalataEast Oilfields in Trinidad, West Indies: in Trans- Advances in Caribbean Geology, Kingston, actions of the Fourth Latin American Geological Con- Jamaica, 18th-20th November, 20-21. ference, Port -of-Spain, Trinidad, 7th-15th July, 1979, 49Persad, K.M. 1989. 1990 Trinidad and Tobago petroleum 2,700-713. consultants encyclopedia. Petroleum Consultants 36Kuarsingh, H.B. 1990. The sedimentology and patrology Trinidad Limited. [sic] of the Herrera Sandstones of Trinidad W.I.: in 50Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of Lame, D.K. & Draper, G. (eds), Transactions of the the Caribbean region; a plate-tectonic perspective: in Twelth Caribbean Geological Conference, St. Croix, Dengo, G. & Case, J.E. (eds), The geology of North U.S. Virgin Islands, 7th-llth August, 1989, p. 551. America, Volume H, The Caribbean region, 405-432. Miami Geological Society, Florida. Geological Society of America, Boulder. 37Kugler, H.G. 1953. Jurassic to Recent sedimentary envi- 51Poole, W.G. 1968. Some sedimentary features of the ronments of Trinidad. Vereinigung Schweizerische Pe- Herrera Sands of the Clarke Road area, Barrackpore troleum-Geologie und Ingenier Bulletin, 20 (59), Oilfield; Trinidad, West Indies: in Saunders, J.B. (ed.), 27-60. Transactions of the Fourth Caribbean Geological Con- 38Kugler, H.G. 1956a. Trinidad—La Trinite with a map: in ference, Port-of- Spain, Trinidad, 28th March-12th Beckmann et aL, Lexique Stratigraphique Interna- April, 1965,101-109. tional, Amerique Latine, Fascicule 2b, Antilles, 41-116, 52Potter, H.C. 1968. A preliminary account of the stratigra- 1 pl. CNRS, Paris. phy and structure of the eastern part of the Northern 39Kugler, H.G. 1956b. Trinidad. Geological Society of Range, Trinidad: in Saunders, J.B. (ed.), Transactions America Memoir, 65,351-365,1 pl. of the Fourth Caribbean Geological Conference, Port - 40Kugler, H.G. 1961. Geological map of Trinidad, scale of -Spain, Trinidad, 28th March-12th April, 1965, 15- 1:100,000. Petroleum Association of Trinidad. 20. 41Kugler, H.G. 1968. Sedimentary volcanism: in 53Potter, H.C. 1973. The overturned anticline of the North- Saunders, J.B. (ed.), Transactions of the Fourth ern Range of Trinidad near Port of Spain. Journal of the Caribbean Geological Conference, Port-of-Spain, Geological Society, London, 129, 133-138. Trinidad, 28th March-12th April, 1965,11-13. 54Potter, H.C. 1976. Type sections of the Maraval, Maracas 42Liska, R.D. 1987. The Cipero and LenguaFormations and Chancellor Formations in the Caribbean Group of and the middle late Miocene boundary: in Duque- the Northern Range of Trinidad: in Cauusse, R. (ed.), Caro, H. (ed.), Transactions of the Tenth/Caribbean Transactions of the Seventh Caribbean Geological Geological Conference, Cartagena de Indias, Conference, Guadeloupe, 30th June-12th July, 1974, Columbia, 14th-22ndAugust, 1983, 204-209. 505-527. B.R.G.M., Orleans. INGEOMINAS, Bogota. 55Potter, H.C. 1985. The geology of the central part of 43Liska, R.D. 1988. The Rio Claro Boulder Bed of central the Northern Range, Trinidad: in Transactions of Trinidad, a sedimentary or tectonic event: in Barker, L. the Fourth Latin American Geological (ed.), Transactions of the Eleventh Caribbean Geologi- Conference, Port-of-Spain, Trinidad, 7th-15th cal Conference, Dover Beach, Barbados, 20th-25th July, 1979, 1, 102-105. July, 1986,12.1-12.7. 56Radovsky, B. & Iqbal, J. 1985. Geology of the North 44Liska, R.D. 1991. The history, age and significance of the Soldado Field: in Transactions of the Fourth Latin Globorotalia menardii Zone in Trinidad and Tobago, American Geological Conference, Port-of- West Indies. Micropaleontology, 37, 173-182. Spain, Trinidad, 7th-15thJuly, 7979,2, 759-769. 45Michelson, J.E. 1976. Miocene deltaic oil habitat, Trini- 57Rambarran, H. 1985. Geology of the Parrylands Field: dad. American Association of Petroleum Geologists in Transactions of the Fourth Latin American Bulletin, 60, 1502-1519. Geological Conference, Port-of-Spain, Trinidad, 46Nath, M. 1985. Production geology of the Coora/Quany 7th-15th July, 1979,2, 770-780. Field: in Transactions of the Fourth Latin American 58Robertson, P. & Burke, K. 1989. Evolution of southern

226

STEPHEN K. DONOVAN

71 Caribbean plate boundary. American Association of Speed, RC. 1987. South American-Antfllean Arc collision Petroleum Geologists Bulletin, 73, 490-509. and the El Pilar Fault: in Duque-Caro, H. (ed.), 59Robertson, P., Burke, K. & Wadge, G. 1986. Structure of Transactions of the Tenth Caribbean Geological Con- the Melajo Clay near Arima, Trinidad and strike-slip ference, Cartagena de Indias, Columbia, 14th-22nd motion in the El Pilar Fault Zone: in August, 1983, p. 170. INGEOMINAS, Bogota. 72 Rodrigues,K.(ed.), Transactions of the First Geological Speed, R.C. & Poland, K.A. 1990. Miocene metamor- Conference of the Geological Society of Trinidad and phism and Neogene tectonics of Northern Range rocks. Tobago, Port-of-Spain, Trinidad, 10th-12th July, Abstracts, Second Geological Conference of the Geo- 1985, 21-33. logical Society of Trinidad and Tobago, Port of Spain, 60Rodrigues, K. 1988. Petroleum source rock evaluation of Trinidad, 2nd-8th April, 19??, 19-20. 73 the Tertiary of Trinidad: in Barker, L. (ed.), Transac- Stainforth, R.M. 1978. Was it the Orinoco? American tions of the Eleventh Caribbean Geological Confer- Association ofPetroleum Geologists Bulletin, 62, 303- ence, Dover Beach, Barbados, 20th-26th July, 1986, 306. 74 38.1-38.16. Suter, H.H. 1960. The general and economic geology of 61Rohr, G.M. 1991. Exploration potential of Trinidad and Trinidad, B. W.I., 2nd edition, with revisionary appendix Tobago. Journal of Petroleum Geology, 14, 343-354. by G.E. Higgins. HMSO, London, 145 pp. 75 62Rowley, K. & Ambeh, W. 1991. The case of the El Pilar Telemaque, C.P. 1990. An evaluation of the geological Fault system in Trinidad and its implications for seis- history and hydrocarbon potential of the late Miocene- mic hazard in the S.E. Caribbean: in Gillezeau, K.A. ?Plio/Pleistocene sediments of the Goudron Field - (ed.), Transactions of the Second Geological Confer- Trinidad: in Lame, D.K. & Draper, G. (eds), Transac- ence of the Geological Society of Trinidad and Tobago, tions of the Twelth Caribbean Geological Conference, St. Port-of-Spain, Trinidad, April 3-8, 1990, 106. Croix, U.S. Virgin Islands, 7th-llth August, 1989, 415- 63Saunders, J.B. 1968. Excursion no. 1 Manzamlla coast: in 429. Miami Geological Society, Florida. 76 Saunders, J.B. (ed.), Transactions of the Fourth Carib- Trechmann, C.T. 1935. Fossils from the Northern Range of bean Geological Conference, Port-of-Spain, Trinidad, Trinidad. Geological Magazine, 72, 166-175. 28th March- 12th April, 1965, 427-429. 77Tyson, L. 1990. Structural features associated with the Los 64Saunders, J.B. 1972. Recent paleontological results from Bajos Fault and their interpretation in the light of current the Northern Range of Trinidad: in Petzall, C. (ed.), theories of strike-slip tectonics: in Larue, D.K. & Draper, Transactions of the Sixth Caribbean Geological Con- G. (eds), Transactions of the Twelth Caribbean ference, Maragarita, Venezuela, 6th-14th July, 1971, Geological Conference, St. Croix, U.S. Virgin Islands, 455-460. 7th-11th August, 1989,403-414. Miami Geological 65Saunders, J.B. 1974. Trinidad: in Spencer, A.M. (ed.), Society, Florida. Mesozoic-Cerrozoic orogenic belts: data for orogenic 78Tyson, L. & Ali, W. 1991. Cretaceous to Middle Miocene studies. Special Publication of the Geological Society, sediments in Trinidad: in Gillezeau, K. A. (ed.), Trans- London, 4, 671-682. [Not seen.] actions of the Second Geological Conference of the 66Saunders, J.B. 1977. Slumping phenomena and olis- Geological Society of Trinidad and Tobago, Port-of- tostromes in the Tertiary geologic history of Trinidad. Spain, Trinidad, April 3-8, 1990, 266-277. GUA Papers on Geology, series 1, no. 9, p. 171. 79Tyson, L., Babb, S. & Dyer, B. 1991. Middle Miocene 67Saunders, J.B. & Bolli, H.M. 1985. Trinidad's contribu- tectonics and its effects on late Miocene sedimentation in tion to world biostratigraphy: in Transactions of the the Trinidad: in Gillezeau, K.A. (ed.), Transactions of the Fourth Latin American Geological Conference, Port- Second Geological Conference of the Geological Society of-Spain, Trinidad, 7th-15th July,1979, 2, 781-795. of Trinidad and Tobago, Port-of -Spain, Trinidad, April 68Saunders, J.B. & Kennedy, J.E. 1968. Sedimentology of a 3-8, 1990, 26-40. section in the Upper Morne 1'Enfer Formation, Guapo 80Vierbuchen, R. 1977. New data relevant to the tectonic Bay, Trinidad: in Saunders, J.B. (ed.), Transactions of history of the El Pilar Fault. GUA Papers on Geology, the Fourth Caribbean Geological Conference, Port-of- series 1, no. 9, 213-214. Spain, Trinidad, 28th March-12th April, 1965, 121- 81Wadge, G. & Macdonald, R 1985. Cretaceous tholeiites of 140. the northern continental margin of South America: the 69Schubert,C. 1979. El Pilar Fault Zone, northeastern Vene- Sans Souci Formation of Trinidad. Journal of the zuela: brief review. Tectonophysics, 52, 447-455. Geological Society, London, 142, 297-308. 70Spath, L.F. 1939. On some Tithonian ammonites from the 82Wall, G.P. & Sawkins, J.G. 1860. Report on the geology of Northern Range of Trinidad. Geological Magazine, 76, Trinidad. Memoir of the Geological Survey, 211 pp. 187-189.

227

Trinidad

[Not seen.] 1991. A preliminary tectonostratigraphic framework 83Wharton, S., Dupigny, A. & Keens-Dumas, J. 1986. A for onshore Trinidad: in Gillezeau, K.A. (ed.), Trans- preliminary investigation of the Brigand Hill Lime- actions of the Second Geological Conference of the stone, Brigand Hill Quarry, Plum Milan Road: in Ro Geological Society of Trinidad and Tobago, Port-of- drigues, K. (ed), Transactions of the First Geological Spain, Trinidad, April 3-8,1990, 41. Conference of the Geological Society of Trinidad and 85Wilson, C.C. 1968. The Los Bajos Fault: in Saunders, J.B. Tobago, Port-of-Spain, Trinidad, 10th-12th July, 1985, (ed.), Transactions of the Fourth Caribbean 102-113. Geological Conference, Port-of-Spain, Trinidad, 28th 84Wielchowsky, C.C., Rahmanian, V.D. & Hardenbol, J. March-12th April, 1965, 87-89.

CHAPTER 13

Northern South America

STEPHEN K. DONOVAN

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION Gibbs and Barron31. The Guyana and Brazilian Shields are major South THE SOUTHERN Caribbean Plate Boundary Zone American plateaux regions separated by the intracratonic (SCPBZ) comprises many independently moving blocks Amazon Basin. In the Guyana Shield older metamorphic and associated basins, and has a total width of circa 250 km, complexes of Archean and early Proterozoic age are over- 70 at least in the southeastern Caribbean . The SCPBZ in- lain by the Roraima Group, comprising relatively unde- cludes Trinidad (see Donovan, chapter 12, this volume), formed clastic sedimentary rocks of middle Proterozoic age. northern Venezuela and northern Colombia (Fig. 13.1). The Older rocks yield abundant evidence of metamorphism, present chapter considers the principal geological and deformation and granite intrusion associated with a Trans- physiographic provinces within continental northern South Amazonian orogeny about 2000 Ma22,31. The Guyana America that are associated with the SCPBZ. The interpre- Shield extends through the three Guyanas (British [=Guy- tation of this zone used herein is broad and includes discus- ana], Dutch [=Suriname] and French) into Venezuela, Co- sion of the Precambrian of the Guyana Shield, that forms the lombia and Brazil (Fig. 13.1). The highest point of the stable craton upon which many of the basins developed and plateau is Mount Roraima (2810 m), famous as "The Lost with which the deformed terranes collided. World” of Sir Arthur Conan Doyle's novel. A detailed This chapter concentrates on the onshore geology of geological map of the Guyana Shield appeared in Gibbs and northern South America, with the exception of the South Barron31. Caribbean Deformed Belt, which occurs both onshore and (mainly) offshore. Discussion of the offshore geology will Archean Imataca Complex be found in contributions on the Caribbean sea floor (Don- nelly, Chapter 3, this volume), and the Netherlands and The Imataca Complex in eastern Venezuela includes the oldest rocks of the Guyana Shield, yielding radiometric Venezuelan Antilles (Jackson and Robinson, Chapter 14, 60,61 this volume). Because the present volume is directed at an dates of 3400 to 3700 Ma for the protolith and compris ing English-speaking audience, I have referred to published quartz-feldspar gneisses associated with granulites and papers in that language. For relevant publications in Span- with localised migmatisation. The principal metamorphic ish, the reader is directed to the comprehensive bibliog- event that affected this complex was the Trans-Amazonian raphies in Case et al.18,20. Further, many facts and orogeny. The parent rocks of the Imataca Complex were interpretations are quoted from review papers, rather than mainly calc -alkaline rocks of continental origin. Other lithologies include intermediate granulites, mafic gneisses, primary sources, as the former are often of the greatest utility 38,39 to the general reader. metasedimentary gneisses, iron formations , ultramafic gneisses, dolomitic marbles and anorthosites31. Granites are present in areas of lower metamorphic grade, at least some STRATIGRAPHY AND STRUCTURE: of which were the product of the Trans-Amazonian orogeny. GUYANA SHIELD The Guri Fault to the south separates the Imataca Complex 18 The following brief introduction to the Guyana Shield is from the rest of the Guyana Shield . based mainly on Read and Watson69, and, particularly,

229 STEPHEN K. DONOVAN

Figure 13.1. The main geological elements of northern South America (after Harrington34, fig. 1; Read and Wat- son69, fig. 9.1; Gibbs and Barron31, fig. 2). Key: AMB=Andean Mobile Belt; CMB=Caribbean Mobile Belt; GS=Guyana Shield; LB=Llanos Pericratonic Basins (Phanerozoic foreland deposits of Andes).

Early Proterozoic granite-greenstone-gneiss belts Greenstone belts dated at about 2250 Ma occur in the do not extend far into the outcrop area of the greenstone north and east of the shield. It is this early Proterozoic belts and rest unconformably on the Trans-Amazonian protolith that gives radiometric ages dating Trans-Ama- succession. Calc -alkaline to alkaline felsic igneous zonian metamorphism. These greenstone belts comprise rocks were erupted in several episodes in the Middle metamorphosed basalts, andesites, dacites and rhyolites, Proterozoic, with mafic dykes and sills intruded between with metagreywackes, phyllites, metaconglomerates, meta- 1700 and 1500 Ma. Middle Proterozoic platform cover morphosed manganiferous and ferruginous sedimentary sequences are undeformed or only locally deformed in rocks, cherts and carbonates 31. Pillow lavas and turbidites association with block faulting and intrusion. indicate accumulation in a submarine environment. The The flat-lying Roraima Group covers an area of 1000 km by 600 km, contains at least a million cubic stratigraphic thickness of the Pastora greens tone belt in 29 Guyana is about 8-10 km, with metamorphosed mafic to kilometres of sediment , and forms flat tablelands with precipitous scarps in Venezuela, Guyana and Brazil. felsic volcanics succeeded by metasedimentary rocks. The metamorphic facies varies from sub-greenschist in the centre The formation includes up to 3700 m of pink and red sandstones, sometimes current bedded, with associated of the belt to amphibolite, produced by thermal metamorphism, 18 near the contact with the Supamo granite18. jasperoid tuffs, conglomerates, siltstones and shales . A The Ile de Cayenne Series in French Guyana is a basal conglomerate is well-exposed in Brazil. Clastics Trans-Amazonian metamorphic complex comprised of me- are arkosic and include detrital diamonds. The environments of deposition were fluvio-deltaic to dium to high metamorphic grade supracrustals, including 31,40 migmatised and granitised metasedimentary rocks of lacustrine . The lowest horizons overlie rhyolitic granulite facies invaded by granitic intrusions. The dominant volcanics in the Guyanas. Basic sills and dykes have fold trends are northeast-southwest and east-west. intruded the formation, and sheets of gabbro and orthonorite reach hundreds of metres in thickness. This Middle Proterozoic intrusion produced doming and gentle folding. Jasperoid tuffs have given radiometric dates of 1730 to The Middle Proterozoic was a time of continental sedi- 1650 Ma. Dolerite intrusions have given radiometric mentation and magmatism. From 1900 to 1500 Ma conti- 18,68 nental sedimentary and volcanic rocks with associated ages of 1600 to 1800 Ma . The source of the sediments granites were formed in the central and western part of the appears to have been the Saharan Shield (Fig. 13.2). shield. These Middle Proterozoic platform cover sequences

230 Northern South America

Figure 13.2. The fit of South America and Africa during the Proterozoic, showing the proximity of the Sahara and Guyana Shields (simplified after Tarling and Tarling81). The arrow shows the direction of transport of the sediments of the Roraima Formation, as indicated by palaeocurrent data, suggesting derivation from the Sahara Shield.

STRATIGRAPHY AND STRUCTURE: NORTHERN VENEZUELA AND COLOMBIA limestones and shales to the north, thinning and becoming sandier to the south. These facies were deposited in deeper The major groupings of basins and uplifts recognised herein and shallower water shelf environments, respectively35. mainly follow those of Case et al18, and I have leaned Less important sources of oil may have been provided by heavily on this seminal paper in preparing the following the Oligocene Merecure Group, the La Pascua and Roblecito account. Figure 13.3A is a map of the region under consid- Formations, and the Miocene Oficina and Chaguaramos eration, showing the geographical relationships of the basins Formations. Oil in these Tertiary rocks is derived from and terranes discussed herein. terrestrial sources. Traps are in Miocene sandstones. Strati- graphic traps are formed by sandstones onlapping onto Basins and arches marginal to the Guyana Shield impervious basement (Fig. 13.4A). Heavy oil plugs are also Eastern Venezuela Basin present. The Eastern Venezuela Basin is an autochthonous se- To the north, traps were originally exploited in Upper quence of up to 15,000 m of Mesozoic and Cenozoic marine Miocene to Pliocene reservoirs, but exploration and exploi- and continental strata resting on Upper Palaeozoic rocks and tation has now extended down into older Tertiary sediments. situated to the north of the Guyana Shield9. The northern These northern fields contain light oil and gas. The most boundary of the basin is marked by a series of northward- important traps are developed in shallow marine sandstones dipping thrusts. Major structural trends are east-west to of Oligocene age, folded into anticlinal structures in the northeast-southwest. The basin is asymmetric, with the northward-dipping thrust belt (Fig. 13.4B). This geological thickest sedimentary pile developed in the north. The basin province underwent a four-phase evolution2: a Cretaceous rests on continental crust. to Eocene passive margin phase, with sediment deposition The Eastern Venezuela Basin is an important petroleum on an open continental shelf; an Eocene to early Miocene province. In the southern part of the basin the so-called foreland basin phase; middle Miocene tectonism as a fore- Orinoco Heavy Oil Belt is the largest oil accumulation land fold and thrust belt; and a subsequent diminished known, with 1200 billion barrels of heavy oil or tar in transpression phase. Post-Eocene deposition was inter- place35. The belt is 460 km by 40 km in area and was rupted by regional orogenic events that produced major yielding 80,000 barrels of oil per day in 1985. Oil is derived unconformities. Traps in the older Tertiary rocks were de- mainly from Upper Cretaceous marine source rocks of the veloped during the Miocene compressional regime. As an Querecual Formation of Cenomanian to Coniacian age. This example of an oil accumulation in this region, the Northern formation is 350-700 m thick, and is comprised of black Monagas field contains possible reserves of 8.6 billion

231 STEPHEN K. DONOVAN

232 Northern South America

barrels trapped in Eocene and Oligocene reservoir sand- For a recent discussion of the basement underlying the stones2. basins discussed above, see Feo-Codecido et al.28. El Baul Swell The El Baul Swell (or Arch or High) is a northwest- Southern Caribbean Plate Boundary southeast trending uplift which separates the Eastern Vene- Zone Falcon Basin zuelan and Barinas-Apure Basins. The El Baul Swell has a The Falcon Basin is a graben structure of the Venezue- core of deformed, slightly metamorphosed Cambro-Ordo- lan borderland whose formation was related to wrench fault- vician shales50 unconformably overlain by faulted Permian ing. Deformed Paleocene to Eocene flysch, associated with or Triassic felsic volcanic rocks. Alkaline plutons of possi- the Siquisiqui terrane, was thrust south during the later bly Palaeozoic antiquity are overlain by Cretaceous and Paleogene. Depos ition of the flysch and of allochthonous Tertiary strata. This horst forms part of the stable South blocks (see below) had ceased by the late Eocene, with uplift 64 American craton. and erosion occurring in the Oligocene to Miocene . Sub- Barinas-Apure Basin sidence related to the wrench fault systems led to the depo- The Barinas -Apure Basin in southwest Venezuela is a sition of up to 6000 m of Neogene clastic sediments and reef sub-Andean basin, southeast of and parallel to the Cordillera limestones. Facies relationships of Oligocene and Miocene 89 de Merida, or Venezuelan Andes. The Guyana Shield forms rocks are complex . Lower Oligocene deltaic sediments are the basement. Locally, Palaeozoic sequences are developed, gradationally overlain by Middle Oligocene to Lower Mio- overlain by up to 4.5 km of mainly marine clastic sedimen- cene marine shales of deeper water origin (1000 m depth), tary rocks of Cretaceous age and up to 5 km of Cenozoic with thin developments of bioclastic and reefal limestones sedimentary rocks. Sedimentary sequences are over 9 km derived as fore-reef debris. These deeper water sediments thick adjacent to the Andes and thin towards the Guyana are in turn overlain by calcareous turbidites, shallow-water Shield21. Older Cenozoic marine clastic and carbonate rocks elastics and reef-related deposits. Locally, basic igneous underlie younger continental deposits. Sedimentary se- sills and dykes of Miocene age intrude Oligocene and Mio- 63,65 quences take the form of five to six coarsening-upward cene sediments in the central part of the basin . Basin megacycles21, with coarser-grained deposits sometimes act- formation occurred under an extensional tectonic regime in 64 ing as reservoirs. Source rocks are Lower Cretaceous shales the Oligocene and Miocene , accompanied by crustal thin- of marine origin. There is moderate folding and extensive ning and basaltic magmatism. Total sedimentary thickness faulting at depth, with hydrocarbon reserves contained in a exceeds 9000 m. Compressive deformation occurred during variety of structural and stratigraphic traps. The Andes have the Miocene and later. The main structural trends are east- been thrust south and east over this basin. The Arauca Arch, west to northeast-southwest, defined by transcurrent and which trends northeast-southwest, separates the Barinas- normal faults, with associated folds and reverse faults. The Apure Basin from the Llanos Basin (Fig. 13.3A) to the underlying crust is transitional or oceanic. This is the only southwest21. northern Venezuelan basin (including offshore sequences)

Figure 13.3A. (Opposite, top) Map of northern South America, showing the relative positions of uplifts and basins discussed herein (simplified after part of Case et al.18, sheet 1; James 35, fig. 1). Key: A=Aruba; B= Bonaire; C=Curacao; G=Grenada; P=Panama; SV=St. Vincent; T=Trinidad: BAB=Barinas-Apure Basin; BGB= Baja Guajira Basin; CB=Cesar Basin; CC=Cordillera Central; CDM=Cordillera de Merida (Venezuelan Andes); CM=Caribbean Mountains; COc=Cordillera Occidental; COr=Cordillera Oriental; CT=Choc6 Terrane; EBS=El Baul Swell; EVB=Eastern Venezuela Basin; FB=Falcon Basin; G=Guajira Peninsula; GS=Guyana Shield; LB=Llanos Basin (not discussed in text); M=Middle Magdalena Basin; MB=Maracaibo Basin; P=Paraguana Peninsula; SCDB=South Caribbean Deformed Belt; SJ=San Jacinto terrane; SJM=San Jorge-Magdalena Prov- ince; SM=Sierra Nevada de Santa Marta; SP=Sierra de Perija: dotted lines=political boundaries; dashed lines= geological boundaries (stratigraphic, unconformable or thrusted); solid lines=faults; coastlines stippled..

Figure 13.3B. (Opposite, bottom) Major faults of northwestern South America (slightly modified after Mann et al.53, fig. 2). Key to fault zones: A=Bocono; B=Santa Marta-Bucaramanga; C=Oca-Chirinos; D=Urdaneta; E=Valera; F=Cuiza; G=Romeral; H=Tigre; I=Cerrejon; J=Avispa; K=Humocaro; L=Otu; M=Cimitarra; N=Palestina; O=North Panama; P=South Caribbean; Q=Moron-El Pilar.

233 STEPHEN K. DONOVAN

Figure 13.4. (A) schematic cross-section across the Eastern Venezuela Basin (simplified after James35, fig. 4). (B) schematic cross-section across the northeast Eastern Venezuela Basin, showing the development of anticlines between imbricate thrusts (simplified after Aymard et al.2, fig. 2). (C) schematic cross-section across the Maracaibo Basin (simplified after James35, fig. 3). Key: K=Cretaceous; E=Eocene; pE=post-Eocene; O=Oligocene; M=Mio- cene; P=Pliocene; u=upper; IR=Interior Range of Caribbean Mountains; HOB=Heavy Oil Belt; SP= Sierra de Perija. Vertical and horizontal scales in km.

234 Northern South America

Figure 13.5. Simplified geological map of the Caribbean Mountains of northern Venezuela (redrawn after Bellizzia and Dengo7, fig. 1). Key: MN=Margarita coastal ophiolite nappes; CCN=Cordillera de la Costa nappe; CTN=Caucagua-El Tinaco nappe (not shown are rocks of pre-Mesozoic age in this belt which may be either autochthonous basement or allochthonous in the nappe); LHN=Loma de Hierro nappe; VCN=Villa de Cura nappe; PN=Piemontina nappe; 'FV' -“Faja Volcada”: EP=E1 Pilar Fault; S=Sebastian Fault: EVB=Eastem Venezuela Basin; FB=Falcon Basin; GP=Gulf of Paria; M=Margarita Island; T=Trinidad.

35 to have so far produced liquid hydrocarbons , although Barquisimeto Trough. Metamorphic grade is greenschist to large discoveries of gas have been made off the Paria Pen- amphibolite or higher, including eclogites, with east-west insula (V. Hunter, written communication). trending foliations. It is uncertain whether there has been The Siquisiqui and Aroa-Misi6n terranes occur in the one or more metamorphic events). Palaeomagnetic evi- south-central and southeast of the Falcon Basin, respec - dence indicates a 90° clockwise rotation of some nappes tively, as elongate, east-west orientated, allochthonous me- since the late Cretaceous8,78. The underlying crust is conti- langes. The Siquisiqui terrane is comprised of gabbros, nental in the south and oceanic in the north. pillow basalts, peridotites, cherts, limestones, shales and The Margarita coastal ophiolite nappe, named after 64 phyllites of Jurassic and Cretaceous age, emplaced in Margarita Island (Fig. 13.5), outcrops as a narrow, discon- Paleogene flysch, and unconformably overlain by Oligo- tinuous coastal belt of metasedimentary and mainly meta- cene to Miocene strata. The Aroa-Mision terrane includes volcanic rocks 7. blocks of metasedimentary and metavolcanic rocks of pre- The Cordillera de la Costa nappe is geologically con- sumed Mesozoic antiquity, with ultramafic rocks and tinuous with the Northern Range of Trinidad (Fig. 13.5). gneissic bodies that may be Precambrian in age, overthrust The pre-Mesozoic basement of the Sebastopol complex is by Paleogene strata. Palaeomagnetic evidence shows that 7 78 composed mainly of metagranites . It is overlain unconfor- these terranes were rotated prior to emplacement . mably by the Caracas Group, a thick sequence of metasedi- Caribbean Mountains mentary rocks with some large granitic intrusions and The Caribbean Mountains consist of deformed, mainly 7 6 metamorphosed mafic igneous rocks . Although formerly allochthonous terranes (Fig. 13.5) that are related geologi- considered to be Mesozoic -Cenozoic, many of the rock units cally to the Mesozoic rocks of Tobago and the Northern within the nappes have now been shown to be Palaeozoic. Range of Trinidad (see Donovan herein, Fig. 12.1). Six For example, the Caracas Group, previously considered to nappes, which have been emplaced south or southeast into be Jurassic to Lower Cretaceous, is intruded by Ordovician Paleogene flysch and which are separated by faulted con- granites (R. Shagam, written communication). Metamor- tacts, are recognised. All nappes have approximately east- phic grade is greenschist to amphibolite facies or higher18. west trends subparallel to the coastline of northern Ave Lallemant1,33 has identified three episodes of deforma- Venezuela (Fig. 13.5), of which only the Piemontina nappe tion of these rocks, in the late Cretaceous, Oligocene to is parautochthonous. The offshore islands of the southern Miocene and the Miocene to Recent. Caribbean Sea form part of the same complex7. The Carib- 5 The Caucagua-El Tinaco nappe is bounded by the La bean Mountains have been complexly metamorphosed and Victoria Fault to the north and the Santa Rosa Fault in the are separated from the Venezuelan Andes to the west by the south. It is constituted of 7:

235 STEPHEN K. DONOVAN

- Discordant volcanic-sedimentary sequence, including the Siquisique ophiolite - Albian to Cenomanian sedimentary forma tions - Tinaquillo peridotites, a mafic-ultramafic complex - El Tinaco complex, the Palaeozoic or older basement

The Loma de Hierro nappe is a narrow zone limited by the Santa Rosa and Agua Fria Faults, with Maastrichtian to Paleocene metasedimentary rocks resting on the Loma de Hierro ophiolite complex7. The Villa de Cura nappe is a sequence of over 5000 m of mainly metavolcanic rocks of blueschist facies18 which represent an ophiolitic assemblage of probable island arc origin7,8. Palaeomagnetic evidence8 indicates 90° clockwise rotation since the Cretaceous. Tectonic emplacement occurred during the late Cretaceous and Paleocene. This nappe rests on continental crust18. The Piemontina nappe or non-metamorphic belt4 on the southern flank of the Caribbean Moun- tains (Fig. 13.5) is an unrotated nappe8 which may be parautochothonous on the South American era- ton. The nappe is bound by the Cantagallo and other faults to the north, and by closely-spaced, northward-dipping thrusts to the south, where the nappe is thrust over the "Faja Volcada" or overturned belt, ______

Figure 13.6. Simplified geological map of the Cordillera de Merida (redrawn after Case et al20, fig. 3). Key: PC/PZ= Precambrian and Palaeozoic metasedimentary and metaigneous rocks; PZ= Palaeozoic sedimentary and metasedimentary rocks (mainly marine); PP= Palaeozoic granitoid plutons; M=Meso- zoic sedimentary and igneous rocks; J=Triassic -Jurassic sedimentary and volcanic rocks (mainly continental); KM=Cretaceous metamorphic rocks; K=Cretaceous sedimentary rocks (mainly marine); T=Tertiary sedimentary rocks (marine and continental): GC=geological contact; F=fault (sense of displacement unknown); TF=thrust fault; SF=strike- slip fault: C=Colombia; V=Venezuela; MB=Maracaibo Basin; SAB=sub- Andean basins.

236 Northern South America

a zone of overturned marine Tertiary rocks at the northern margin of the Eastern Venezuela Basin. The highly-folded sequence of the Piemontina nappe is as follows7:

= 2000 m Maasthchtian to flysch early Eocene

= 1500 m Campanian to siliceous shales Maastrichtian and limestones

= 2000 m Coniacian flysch

= 250-500 m Cenomanian shales and Turonian limestones

Deformed uplifts with Precambrian to Palaeozoic cores The following geologic provinces within the Colombian region comprise Precambrian and Pa- laeozoic metamorphic and igneous rocks overlain by younger cover sequences. Cordillera de Merida The Cordillera de Merida, or Venezuelan An- des, is a vertically-uplifted, seismically-active, structural block (Fig. 13.6). The basement consists of Precambrian(?) (possibly an outlier of the Guyana Shield) and Palaeozoic metamorphic and igneous rocks. These, in turn, are overlain by Palaeozoic marine sedimentary rocks and Upper Carboniferous to Permian red beds. Marine Lower Palaeozoic rocks in the southern piedmont are unmetamorphosed, while Upper Palaeozoic flysch in the central ranges has undergone regional metamorphism. Unconformably overlying these Palae- ozoic rocks are Triassic(?)-Jurassic continental sedimentary and volcanic rocks, Cretaceous marine clastic and carbonate, Paleogene marine, and Neogene continental clastic sedimentary rocks20. The total thickness of Phanerozoic strata is greater than 10 km. Emplacement of granitoid plutons occurred in the late Precambrian(?), late Cambrian to early Ordovician, Devonian and Permian to early Triassic 20. The block has been complexly ______

Figure 13.7. Simplified geological map of the Cordillera Oriental (redrawn after Case et al.20, fig. 7). The complex association of different sequences in the northern half of the map comprises the Santander massif. Key: PC=Precambrian metamorphic and igneous rocks; PC/PZ=Precambrian and Palaeozoic metamorphic and igneous rocks; PM=Palaeozoic metamorphic rocks; PP=Palaeozoic granitoid plutons; PZ=mainly marine Palaeozoic sedimentary rocks; M=Mesozoic sedimentary and volcanic rocks; MP=Mesozoic granitoid plutons; JV= Jurassic volcanic rocks; J=mainly continental Jurassic sedimentary and volcanic rocks; K=mainly marine Cretaceous sedimentary rocks; T=mainly continental Cenozoic sedimentary rocks: CM=Cordillera de Merida; SP= Sierra de Perija. Key to geological boundaries as for Figure 13.6.

237 STEPHEN K. DONOVAN

Figure 13.8. Simplified geological map of the Cordillera Central (CC), the Cordillera Occidental (CO) and the Romeral Fault Zone (RFZ), redrawn after Case et al.20. Key: PC=Precambrian igneous and metamorphic rocks; PM=Palaeozoic metamorphic rocks; PP=Palaeozoic grantoid plutons; MP=Mesozoic and Palaeozoic granitoid plutons; J=Mesozoic (mostly Jurassic) volcanic and continental sedimentary rocks; KV=Cretaceous mafic and intermediate subaqueous volcanic rocks; U=mainly intrusive mafic and ultramafic rocks (mostly Cretaceous); KM=Cretaceous metamorphic rocks; K=Cretaceous sedimentary rocks; TP=Tertiary granitoid plutons; TV= Neogene and Quaternary volcanic rocks; T=Paleogene (mainly marine) and younger (mainly continental) sedimentary rocks; AB=Antioquian Batholith. Key to geological boundaries as in Figure 13.6.

238 Northern South America

deformed by numerous Phanerozoic events, the four princi- Jurassic -Cretaceous siliciclastics and local pal orogenic episodes occurring in the late Precambrian(?), carbonates Devonian to early Carboniferous, late Permian (associated with high grade regional metamorphism) and end Eo- Devonian-Permian or 2000-2500 m of clastics cene73,74. Late Cenozoic uplift was complex and Triassic and carbonates, with red polyphase43. Younger rocks have a generally northeast- beds towards the top of the southwest structural trend20, including the active, axial Bo- sequence cono Fault Zone (Fig. 13.3B). Movement on this fault is predominantly dip-slip with lesser strike-slip71,72 (R. Sha------unconformity------gam, written communication) and it is interpreted by some as part of the SCPBZ, although all of the movement between Cambrian-Ordovician Silgar and Quetame the two plates cannot be taken up on this structure alone32. Series —metamorphic rocks of The faults flanking the Cordillera de Merida are thrusts and greenschist to lower reverse faults that dip towards the core of the uplift20. The amphibolite facies Cordillera de Merida is separated from the Cordillera Ori------unconformity------ental by the Tachira depression, which represents a zone of thrusting towards the southwest52. Precambrian Bucaramanga Gneiss— Cordillera Oriental metamorphic rocks of high The core of the Cordillera Oriental (Fig. 13.7) exposes amphibolite facies Precambrian to Palaeozoic metamorphic and minor igneous rocks. Precambrian gneisses in the southern part of the 20 province give radiometric ages of 1600 Ma . These are However, this stratigraphic scheme is probably over- overlain by Palaeozoic sedimentary rocks, Mesozoic conti- simplified. For example, some formation boundaries corre- nental deposits, volcanics, evaporites and marine elastics, spond to metamorphic isograds and the Silgar Series with thick, mostly continental Cenozoic successions devel- metawackes have been traced laterally into Bucaramanga 14,37 oped locally . There are more than 10,000 m of post-Pa- Gneiss (R. Shagam, written communication). Conceivably 74 laeozoic strata . In the southern part of the range, some of the crystalline rocks have no stratigraphic signifi- 20,37 Cretaceous(?) evaporites have produced salt anticlines . cance, but are merely different metamorphic facies of the Principal deformation episodes occurred during the Precam- same rock unit.Granite plutonism occurred in the late Pre- brian, Palaeozoic and late Cretaceous to Paleogene, with cambrian and late Ordovician, with the main phase of batho- uplift following in the post-Miocene. The main structural lith emplacement between the late Triassic and late trend is north-south. Flanking thrust faults dip beneath the Jurassic20. range, although most faults are high-angle structures. The Sierra de Perija 20 principal fault, the Santa Marta-Bucaramanga Fault (Fig. The Sierra de Perija uplift is similar in both stratigraphy 13.3B), is seismically active. It is believed by many to be a and structure to the Cordillera Oriental and the Venezuelan left-lateral strike-slip fault and by others to be an oblique Andes. The underlying crust is continental. Precambrian(?) slip high-angle thrust. Folds are broad and open in the north, and Palaeozoic crystalline rocks are sparsely exposed on the but complex in the south where they are related to salt eastern flanks20. Palaeozoic marine and continental?) strata tectonics. This range was intruded by granitoid plutons in occur as isolated outcrops, with Palaeozoic and Mesozoic the Palaeozoic and Mesozoic, mainly in the area of the volcanic, plutonic and continental sedimentary rocks con- Santander Massif in the northern part of the province, and centrated in the core of the uplift. Mesozoic red beds, has been autochthonous on South America since the Juras- 49,50,75 associated with hypabyssal intrusions and interbedded with sic . Regional metamorphic events occurred in the volcanic rocks, are best developed in northnortheast-south- late Precambrian, late Ordovician to late Silurian, and pos - 55,79 74 southwest trending extensional grabens . Jurassic and sibly early Carboniferous and Permian . The underlying Cretaceous marine strata are exposed at high elevations. At crust is continental. least eight major deformation episodes have affected the The Santander Massif (Fig. 13.7) in the north of the Sierra de Perija during the Phanerozoic, with four in the Cordillera Oriental is one of three major uplifts within the Cenozoic, in the early Eocene, middle Eocene, late Oligo- range37. In the massif the following sequence of rocks have 42 20,74 cene and Pliocene . Major structural trends are northeast- been mapped : southwest. This range was uplifted in the post- Miocene20,41,42 and is still seismically active. Available palaeomagnetic data from Jurassic rocks suggests that the Sierra de Perija and the Cordillera de Merida had different patterns of tectonic movement during the Mesozoic 20,56.

239

STEPHEN K. DONOVAN

Figure 13.9. The geological evolution of northern South America during the Cenozoic (after Dewey and Pindell24, fig. 2; Pindell and Barrett66, fig. 10). A, Paleocene. B, Eocene. C, Oligocene. D, early Miocene. E, late Miocene. F, Holocene. For explanation see text. Key: heavy stipple=continental crust overthrust by oceanic terranes; light stipple=flysch of the Scotland Formation, Barbados; VVV=island arc volcanism; heavy arrows= relative motion of South American and Caribbean Plates; anticlinal symbols=forehand flexural bulge of advancing thrust load of oceanic terranes.

Cordillera Central flanking sedimentary rocks are developed, which have been The core of the Cordillera Central (Fig. 13.8) uplift is moderately folded and strongly faulted. The rock associa- comprised of Precambrian to (mainly?) Palaeozoic igneous tion of the Cordillera Central constitutes a polydeformed and metamorphic rocks, with gneisses of granulite facies metamorphic complex of Precambrian to Cretaceous unconformably overlain by fossiliferous Ordovician me- rocks 82. Large Tertiary and (mainly) Mesozoic plutons have tasedimentary rocks of low greenschist facies20. The oldest intruded this province. The great Antioquian Batholith (Fig. radiometric dates obtained are in excess of 1300 Ma82. The 13.8) of the centre of the Cordillera was mainly intruded Palaeozoic island arc-related sequence was metamorphosed during the Cretaceous into high-grade metamorphic rocks in the late Devonian to early Carboniferous58,59. This core of uncertain (Palaeozoic?) age, which had previously been is overlain in the east by Triassic limestones and Jurassic intruded by Palaeozoic granitoid plutons that were them- continental deposits and calc-alkaline volcanic rocks, suc- selves subsequently metamorphosed20. Ultramafic and ceeded by Cretaceous marine clastic and carbonate sedi- mafic bodies, possible fragments of upper mantle or lower mentary rocks20. In the west obducted20 oceanic and mantle crust, outcrop in the north and west. The main structural rocks occur in a thrust zone. Up to 4000 m of Mesozoic trends are north-south to northeast-southwest, most large

240 Northern South America

faults having a major strike-slip component, such as the major fault separates the north and south sections of the right-lateral PalestinaFault 27 (Fig. 13.3B). The eastern mar- province. Tertiary basins rest unconformably on these older gin of the cordillera is bounded by a narrow, east-verging strata and contain a total thickness of at least 4000 m of fold and thrust belt12,20. The underlying crust is continental marine clastic and carbonate rocks, with some continental and, in obducted slices, oceanic20. strata20. The terrane has been intruded by felsic plutons19,20. The northern limit of active Colombian volcanism oc - The crust is oceanic, transitional or continental, and is not curs in the southern part of the Cordillera Central, related to in isostatic equilibrium (cf. Sierra Nevada de Santa Marta). plate convergence at the Colombia Trench, and thick Neo Palaeomagnetic evidence indicates rotation of Cretaceous gene volcanic terraces are developed. Cenozoic volcanism rocks in the southern part of the province20,49 . in this region commenced in the late Oligocene to early Miocene84 or late Miocene54, with the formation of calc-al- Mesozoic-Cenozoic deformed belts kaline stratovolcanoes erupting mainly andesitic lavas. Re- Cordillera Occidental cent eruptions include that of Nevado del Ruiz on 13th The Cordillera Occidental (Fig. 13.8) rests on oceanic November, 1985, which generated lahars that flowed down crust17. It is a complex, polycomponent terrane20 comprised of flanking river valleys and killed between 22,000 and 25,000 a great thickness of Mesozoic eugeosynclinal rocks17 people57, the fourth highest recorded death toll from a including Cretaceous tholeiitic basalts and basaltic an- volcanic event. desites, with sedimentary rocks of pelagic and turbiditic Sierra Nevada de Santa Mart a origin, and local developments of ultramafic rocks 30, all of The Sierra Nevada de Santa Marta massif is a pyrami- which were intruded by Tertiary granitoid plutons. The dal, poly component uplift with a northeast-southwest struc- principal structural trend is north-south. Beds are often tural trend that defines three parallel metamorphic steep, vertical or overturned17. Major fault systems border terranes 83. It is the highest mountain range in Colombia, the Cordillera Occidental in the east and west20. The major reaching a maximum elevation of about 5800 m42. It is deformation episode of this province occurred before the bordered to the north by the Oca Fault (right slip) and to the mid-Miocene20, but great vertical uplift has continued into west by the Santa Marta Fault (left slip), both with large the Recent. vertical components (Fig. 13.3B). Precambrian gneisses in The Romeral Fault Zone (Figs 13.3B, 13.8), of late the core of the uplift have given radiometric Rb-Sr dates of Cretaceous to Paleogene antiquity, forms a suture zone circa 1300 Ma20,48. The core is overlain by some Palaeozoic between the Cordillera Occidental, on oceanic crust, and the sedimentary rocks, and Mesozoic volcanics and volcani- Cordillera Central, on continental crust17. Principal struc - clastic rocks, and was intruded by Phanerozoic plutons. The tural trends are north-south to northeast-southwest, but Santa Marta schists form a Mesozoic subterrane to the whether the terrane is dipping towards the east or west is northwest47, related to similar sequences in the Guajira uncertain20. The sequence comprises a 'megamelange' of Peninsula (the Ruma metamorphic belt), in Aruba, the Para- ultramafic rocks, ophiolite fragments, scattered blueschists, guan Peninsula and the Cordillera de la Costa of the Carib- fragments of continental crust, and Mesozoic flysch and bean Mountains 20. The crust is continental, but geophysical pelagic sedimentary rocks18. A superimposed Tertiary ba- evidence shows that it si not in isostatic equilibrium10,19. sin, infilling a graben-like structure17, includes circa 3000 The Santa Marta Massif was uplifted and transported north- m of mainly continental sedimentary and volcanic rocks westwards over adjacent basins on a low angle thrust fault, which are locally tightly folded20. Tertiary igneous rocks and between the Oca and Santa Marta Faults, during the have been rotated anticlockwise46. Pliocene42. Palaeomagnetic analysis of Jurassic strata indi- Choco Terrane cates no large tectonic rotations have occurred since the The Choco Terrane (Fig. 13.3A) is comprised of up- mid-Mesozoic, in contrast to those recognised for rocks of lifted oceanic crust and magmatic arc rocks of late Creta- the Guajira Peninsula to the north51. ceous to Paleogene age, which adjoin deep forearc basins Guajira Peninsula containing up to 10,000 m of pelagic, turbiditic andmarginal Precambrian and Palaeozoic metamorphic rocks out- marine sedimentary strata of Cretaceous and Cenozoic age. crop in the core of the Guajira Peninsula This core is To the west, the Choco Terrane is bordered by a partly-in- overlain by a thick sequence of Mesozoic clastic and car- filled trench and is overriding the Nazca Plate26. The main bonate sedimentary rocks, with local volcanic rocks, all of structural trends are north-south, northeast-southwest and which are highly deformed. The Ruma metamorphic belt northwest-southeast. The Choco Terrane is overridden to (related to the Santa Marta schists; see above) in the north the east by the Cordillera Occidental and the South Carib- is composed of Mesozoic metamorphic and ultramafic rocks bean Deformed Belt along the Atrato Fault, an eastward-dip- related to end Cretaceous to Paleocene subduction20. A ping reverse fault18,26. The Choco Terrane continues

241 STEPHEN K. DONOVAN

northwards into eastern Panama15,26. crust is oceanic or undefined. San Jacinto Terrane The San Jacinto Terrane extends to the north of the Andean basins on continental crust Cordillera Occidental and is probably underlain by oceanic Maracaibo Basin crust. It is constituted of late Cretaceous to Neogene marine The Maracaibo Basin is an Andean basin, sandwiched pelagic, turbiditic, clastic and carbonate sedimentary rocks, between the Venezuelan Andes and the Sierra de Per ija. This with associated fluvial and lacustrine deposits, to a total basin has a total thickness of about 11,000 m and has been thickness of up to 10,000 m. Cretaceous strata have been the most important of the three major petroleum provinces intruded by tonalite plutons. The principal structural trend is in Venezuela35. It includes continental deposits of Jurassic northnortheast-southsouthwest, with Cenozoic deforma- age, at least 1000 m of Cretaceous marine carbonates and tional events and Paleogene diapirism having produced clastic sedimentary rocks, and 5000-7000 m of Tertiary strong deformation18,20, with broad, gentle synclines separated deltaic, fluvial, lacustrine and marine deposits (Fig. 13.4C). by narrow, elongate anticlines25. Hydrocarbons are comprised of heavy oils in shallow reser- South Caribbean Deformed Belt voirs, with lighter oil and gas at depth. Production is from The South Caribbean Deformed Belt (SCDB; Fig. stratigraphic traps in shallow Tertiary sandstones (north- 13.3A), with its eastern extension as the Curacao east88), or structural traps in Tertiary sandstones and Creta- Ridge16,77, is situated offshore Colombia and Venezuela and ceous limestones (central and western lake), or Tertiary rests on oceanic crust. It consists of up to 10,000 m of sandstones, Cretaceous limestones and (locally) base- poorly consolidated and highly deformed Cretaceous and ment35. The principal source rock is the Cenomanian-San- Cenozoic sedimentary rocks and sediments45 of pelagic and tonian La Luna Formation, an organic -rich, dark grey to turbiditic origin, derived from South America and deposited black limestone to calcareous shale with cherts and of ma- on the Caribbean Plate during its eastward drift23. rine origin. It is uncertain whether the La Luna Formation Northeast-southwest and east-west structural trends (folds is deep- or (more probably) shallow-water in origin, but it and thrust faults) are related to the oblique southeastward overlies the (mainly) shallow-water marine limestones of convergence of the Caribbean Plate with northern South the Cogollo Group3. Production is mainly from the north, America36,44 at an estimated rate of 20 mm per year. This but in the south the North Andean Foredeep also contains convergence has produced the deformation of the SCDB and important reserves35. Thick sedimentary sequences in the induced uplift of the Venezuelan Andes, Sierra de Perija, southern part of the basin (the late Cenozoic Guayabo delta) Guajira Penin sula, Netherlands Antilles, northwest Caribbean include near-shore to non-marine sequences produced by Mountains and the Sierra Nevada de Santa Marta. Mud erosion from the Cordillera Oriental and, subsequently, the diapirism is widespread in the west of this terrane25,87. Cordillera de Merida85. Active deformation is probably concentrated along the The basement is metamorphosed continental crust. northern margin of the SCDB, with the effects of Faulting is complex and increases with depth, with moderate deformation becoming progressively older towards the folding adjacent to faults. The Oca Fault (Fig. 13.3B) is a south45, with slices of oceanic crust possibly incorporated into major east-west, dextral strike-slip or oblique slip structure the belt. In western Colombia, the SCDB is concealed by the that bounds the north of the basin86. The main structural northward-prograding, undeformed sediment apron of the trends are north-south and northeast-southwest. Magdalena Fan11,76. The SCDB comes ashore as the Sinu Middle Magdalena Basin Terrane in northwest Colombia25. The Middle Magdalena Basin is an intermontane basin Paraguana Peninsula situated between the Cordillera Oriental and the Cordillera The Paraguan Terrane is an uplifted block in northwestern Central. This basin contains at least 6000 m of Mesozoic Venezuela with the following rock sequence18: marine14 and continental sedimentary (and volcanic?) rocks, and Cenozoic marine and, mainly, continental depos- 18,20 Tertiary = 3000 m of gently deformed strata its . Marine deposition largely ceased after the Creta- ceous following the uplift of the Cordillera Central20. Mesozoic metavolcanic and metasedimentary Neogene volcaniclastics in the west were derived from rocks, including an anorthositic stratovolcanoes in the Cordillera Central84. The Cenozoic gabbro and an ultramafic complex sedimentary sequence thickens eastwards14. Structural Palaeozoic metaigneous rocks trends are north-south to northeast-southwest. Deformation (dominantly faults and associated minor folds, with associated local unconformities) increases eastwards14, and south- Palaeomagnetic evidence indicates that this terrane has 20 been rotated during regional deformation78. The underlying wards into the Upper Magdalena Basin , and folds increase

242 Northern South America

in abundance northwards13. The Cordillera Oriental is thrust Cretaceous to Paleocene over the Middle Magdalena Basin. This basin is an impor- The northern Venezuelan nappes were emplaced onto tant petroleum province. The underlying crust is continental. the margin of the South American Plate diachronously, San Jorge-Magdalena Basin younging from west to east, between the Cretaceous and The San Jorge-Magdalena province is between the Cor- Pliocene. This pattern of emplacement-was closely related dillera Central and the Sierra Nevada de Santa Marta. The to the eastward progression of the Caribbean Plate during crystalline basement is comprised of Precambrian and Pa- the Cenozoic. Nappe emplacement and the development of laeozoic rocks similar to those of the Cordillera Central. an associated foredeep occurred in the Maracaibo Basin 3000 to 8000 m of overlying (mainly) Tertiary continental during the Paleocene to early Eocene (Fig. 13.9A), with the and marine strata are moderately deformed, with a north- accumulation of turbidites within the basin and the offshore east-southwest structural trend. The basin is an important trench; in the Caribbean Mountains in the Oligocene to hydrocarbon province18,20. Miocene (Fig. 13.9B-D); and in the Eastern Venezuela Cesar Basin Basin in the late Miocene to Pliocene (Fig. 13.9E, F). The The Cesar Basin is located between the Sierra Nevada Grenada Basin also opened in the Paleocene due to intra-arc 66 de Santa Marta and the Sierra de Perija in northeastern spreading . Colombia. The asymmetrical basin fill thickens southeast13, Eocene and Oligocene and is comprised of up to 1000 m of Mesozoic strata overlain The presumed offset of the Caribbean and South by a similar thickness of marine and (mainly) continental American Plates in the late Eocene was 1400 km80. East- sediments of Cenozoic age, with major lignite and coal ward-dipping subduction of the Caribbean Plate at the Sinu deposits in the northeast18,20. The principal structural Trench commenced in the Eocene25. The eastward advance of components are broad folds and high-angle faults trending the Caribbean Plate produced diachronous accretion of northeast-southwest. The Sierra de Perija may be thrust turbidites. Nappes and the turbiditic accretionary complex over this basin20,42. The underlying crust is continental18,20. were emplaced onto the northern South American margin, 66 The stratigraphy and structure of the Cesar and Middle with associated dextral shear and the clockwise rotation of 8,78 MagdalenaBasins are similar, and Campbell has proposed rock assemblages within nappes . The Oligocene sub- that the two were continuous with each other before offset- sidence of the Falcon Basin originated by continental ting by the northnorthwest-southsoutheast trending Santa stretching and pull-apart basin development due to dextral Marta Fault (for an alternative view, see Poison and strike-slip motions. Henao67; Fig. 13.3B). Campbell's suggestion has been sup- Early to late Miocene ported by more recent data and analyses42. The commencement of the Andean orogeny during the Baja Guajira Basin Miocene led to the northeastward migration of Andean 53 The Baja Guajira Basin is situated between the Oca terranes, such as the movement of the Maracaibo Block Fault to the south and the Guajira Province to the north (Fig. along the Bocono and Santa Marta-Bucaramanga Fault 13.3A). A northwest-southeast trending concealed fault Zones (Fig. 13.9D-F). Thus, basins and terranes within the may border the northern margin of the basin19. The basin South Caribbean Plate Boundary Zone have migrated north- 23 fill consists of up to 4000 m of Eocene(?) to Pliocene clastic east and east with respect to the South American craton . sedimentary rocks of both terrestrial and marine origin18. This resulted in the Caribbean/South American transform Deformation is moderate, and is probably related to pull- fault becoming convergent, with the Venezuelan borderland apart movements on the east-west trending Cuiza (to the overriding the Caribbean Plate. north) and OcaFaults19,20. The underlying crust is continental, Present at least towards the south. The Panama arc of Central America collided with the Western Cordillera along the Atrato Suture during the Plio- 23 CENOZOIC GEOLOGICAL EVOLUTION cene and Pleistocene . The topography of the Andes has been produced by continued convergence of these terranes. The model for the evolution of the South Caribbean Plate The Netherlands and Venezuelan Antilles were obducted Boundary Zone followed herein (Fig. 13.9) is that proposed onto the Caribbean Plate during the Neogene or earlier, by Dewey and Pindell23,24,66. Although other models for the associated with the development of the South Caribbean evolution of the Caribbean Plate exist (see, for example, Deformed Belt. Morris et al.62), the Dewey and Pindell model is considered ACKNOWLEDGMENTS—I thank the American Geophysical Union, Jim the most feasible paradigm available. The following account 66 Pindell and John Dewey for granting permission to reproduce Figure 13.9. This is based mainly on that of Pindell and Barrett . paper was greatly improved in the light of incisive review comments by William B. MacDonald, Vernon Hunter and, particularly, Reginald

243 STEPHEN K. DONOVAN

10 Shagam. Bonini, W.E, Garing, J.D. & Kellogg, J.N. 1982. Late Cenozoic uplifts on the Maracaibo-Santa Marta block, slow subduction of the Caribbean Plate, and results REFERENCES from a gravity study: in Snow, W, Gil, N, Llinas, R, Rodriguez-Torres, R.; Seaward, M. & Tavares, I. (eds), 1Ave Lallemant, H.G. 1990. The Caribbean-South Ameri- Transactions of the Ninth Caribbean Geological Con- can plate boundary, Araya Peninsula, eastern Vene- ference, Santo Domingo, Dominican Republic, 16th- zuela: in Larue, D.K. & Draper, G. (eds), Transactions 20th August, 1980,1, 99-105. 11 of the Twelfth Caribbean Geological Conference, St. Breen, N.A. 1989. Structural effect of Magdalena Fan Croix, U.S. Virgin Islands, 7th-llth August, 1989,461- deposition on the northern Colombia convergent mar- 471. Miami Geological Society, Miami. gin. Geology, 17, 34-37. 2Aymard, R., Pimentel, L., Eitz, P., Lopez, P., Chaouch, A., 12Butier, K. & Schamel, S. 1988. Structure along the eastern Navarro, J., Mijares, J. & Pereira, J.G. 1990. Geological margin of the Central Cordillera, Upper Magdalena integration and evaluation of Northern Monagas, East- Valley, Colombia Journal of South American Earth ern Venezuela Basin: in Brooks, J. (ed.), Classic Petro- Sciences, 1,109-120. leum Provinces. Geological Society Special 13Campbell, C.J. 1968. The Santa Marta wrench fault of Publication, 50, 37-53. Colombia and its regional setting: in Saunders, J.B. 3Bartok, P., Reijers, T.J.A. & Juhasz, 1.1981. Lower Creta- (ed.), Transactions of the Fourth Caribbean Geological ceous Cogollo Group, Maracaibo Basin, Venezuela: Conference, Port-of- Spain, Trinidad, 28th March-nth sedimentology, diagenesis, and petrophysics. Ameri- April, 1965, 241-261. can Association of Petroleum Geologists Bulletin, 65, 14Campbell, C.J. & Burgl, H. 1965. Section through the 1110-1134. Eastern Cordillera of Colombia, South America. Geo- 4Beck, C.M. 1978. Polyphasic Tertiary tectonics of the logical Society of America Bulletin, 76, 567-590. interior range in the central part of the western Carib- 15Case, J.E. 1974. Oceanic crust forms basement of eastern bean chain, Guarico state, northern Venezuela Geolo- Panama. Geological Society of America Bulletin, 85, gieenMijnbouw, 57, 99-104. 645-652. 5Beets, D.J., Maresch, W.V., Klaver, G.T., Mottana, A., 16Case, J.E. 1974. Major basins along the continental mar- Bocchio, R., Beunk, F.F. & Monen, H.P. 1984. Mag- gin of northern South America: in Burk, C.A. & Drake, matic rock series and high-pressure metamorphism as C.L. (eds), The Geology of Continental Margins, 733- constraints on the tectonic history of the southern Car - 741. Springer-Verlag, New York. ibbean. Geological Society of America Memoir, 162, 17Case, J.E., Duran S., L.G., Lopez R., A. & Moore, W.R. 95-130. 1971. Tectonic investigations in western Colombia and 6 Bellizzia, A. 1972. Is the entire Caribbean Mountain belt eastern Panama. Geological Society of America Bulle- of northern Venezuela allochthonous? Geological So- tin, 82, 2685-2712. ciety of America Memoir, 132, 363-368. 18Case, J.E., Holcombe, T.L. & Martin, R.G. 1984. Map of 7 Bellizzia, A. & Dengo, G. 1990. The Caribbean Mountain the geologic provinces in the Caribbean region. Geo- system, northern South America; a summary: in Dengo, logical Society of America Memoir, 162,1-30. G. & Case, J.E. (eds), The Geology of North America. 19Case, J.E. & MacDonald, W.D. 1973. Regional gravity Volume H. The Caribbean Region, 167-175. Geologi- anomalies and crustal structure in northern Colombia. cal Society of America, Boulder. Geological Society of America Bulletin, 84, 2905-2916. 8 Blin, B., Sichler, B. & Stephan, J.F. 1990. Contribution of 20Case, J.E., Shagam, R. & Giegengack, R.F. 1990. Geology paleomagnetism in the relations between the Piemontin of the northern Andes; an overview: in Dengo, G. & and internal nappes of the Venezuelan Caribbean chain: Case, J.E. (eds), The Geology of North America. Vol- synthesis and new data: in Larue, D.K. & Draper, G. ume H. The Caribbean Region, 177-200. Geological (eds), Transactions of the Twelfth Caribbean Geologi- Society of America, Boulder. cal Conference, St. Croix, U.S. Virgin Islands, 7th-llth 21Chigne, N. & Hernandez, L. 1990. Main aspects of petro- August, 1989,453-460. Miami Geological Society, Mi- leum exploration in the Apure area of southwestern ami. Venezuela, 1985-1987: in Brooks, J. (ed.), Classic Pe- 9 Bonini, W.E. 1978. Anomalous crust in the Eastern Vene- troleum Provinces. Geological Society Special Publi- zuela Basin and the Bouguer gravity anomaly field of cation, 50, 55-75. northern Venezuela and the Caribbean borderland. 22Cordani,U.G., Melcher, G.C. & de Almeida, F.F.M. 1968. Geologie en Mijnbouw, 57, 117-122. Outline of Precambrian geochronology of South Amer-

244 Northern South America

ica Canadian Journal of Earth Sciences, 5, 629-632. Society of America Bulletin, 81, 3623-3646. 23Dewey, J.F. & Pindell, J.L. 1985. Neogene block tectonics 38Kalliokoski, J. 1965. The metamorphosed iron ore of El of eastern Turkey and northern South America: conti- Pao, Venezuela. Economic Geology, 60, 100-116. nental applications of the finite difference method. Tec- 39Kalliokoski, J. 1965. Geology of north-central Guayana tonics, 4, 71-83. Shield, Venezuela Geological Society of America 24Dewey, J.F. & Pindell, J.L. 1986. Neogene block tectonics Bulletin, 76, 1027-1050. of eastern Turkey and northern South America: conti- 40Keats, W. 1976. The Roraima Formation in Guyana: a nental applications of the finite difference method: revised stratigraphy and a proposed environment of reply. Tectonics, 5, 703-705. deposition: in Boletin de Geologia Publicacion Espe- 25Duque-Caro, H. 1984. Structural style, diapirism, and cial, 1. Memoria, Segundo Congreso Latinoamericano accretionary episodes of the Sinu-San Jacinto terrane, de Geologia, Caracas, Venezuela, 11 al 16 de southwestern Colombia borderland. Geological Soci- Noviembre de 1973,2,901-940. ety of America Memoir, 162, 303-316. 41Kellogg, J.N. 1982. Cenozoic basement tectonics of the 26Escalante, G. 1990. The geology of southern Central Sierra de Perija, Venezuela and Colombia: in Snow, W., America and western Colombia: in Dengo, H. & Case, Gil, N., Llinas, R., Rodriguez-Torres, R, Seaward, M. I.E. (eds), The Geology of North America. Volume H. & Tavares, I. (eds), Transactions of the Ninth Carib- The Caribbean Region, 201-230. Geological Society of bean Geological Conference, Santo Domingo, Domini- America, Boulder. can Republic, 16th-20th August, 1980,1 107-117. 27Feininger, T. 1970. The Palestina Fault, Colombia Geo- 42Kellogg, J.N. 1984. Cenozoic tectonic history of the Sierra logical Society of America Bulletin, 81,1201-1216. de Perija, Venezuela-Colombia, and adjacent basins. 28Feo-Codecido, G., Smith, F.D., Jr., Aboud, N. & Gia- Geological Society of America Memoir, 162,239-261. como, E. de D. 1984. Basement and Paleozoic rocks of 43Kohn, B.P., Shagam, R, Banks, P.O. & Burkley, L.A. the Venezuelan Llanos basins. Geological Society of 1984. Mesozoic -Pleistocene fission-track ages on rocks America Memoir, 162, 175-187. of the Venezuelan Andes and their tectonic implica- 29Gansser, A. 1954. The Guiana Shield (S. America): geo- tions. Geological Society of America Memoir, 162, logical observations. Eclogae Geologicae Helvetiae, 365-384. 47, 77-112. 44Ladd, J.W., Holcombe, T.L., Westbrook, O.K. & Edgar, 30Gansser, A. 1973. Facts and theories on the Andes. Jour- N.T. 1990. Caribbean marine geology; active margins nal of the Geological Society, London, 129, 93-131. of the plate boundary: in Dengo, G. & Case, J.E. (eds), 31Gibbs, A.K. & Barron, C.N. 1983. The Guiana Shield The Geology of North America. Volume H. The Carib- reviewed. Episodes, 6 (2), 7-14. bean Region, 261-290. Geological Society of America, 32Giegengack, R. 1984. Late Cenozoic tectonic environ- Boulder. ments of the central Venezuelan Andes. Geological 45Ladd, J.W., Truchan, M., Talwani, M., Stoffa, P.L., Buhl, Society of America Memoir, 162, 343-364. P., Houtz, R, Mauffret, A. & Westbrook, G. 1984. 33Guth, L.R. & Ave Lallemant, H.G. 1990. A kinematic Seismic reflection profiles across the southern margin history for eastern Margarita Island, Venezuela: in of the Caribbean. Geological Society of America Mem- Lame, D.K. & Draper, G. (eds), Transactions of the oir, 162, 153-159. Twelfth Caribbean Geological Conference, St. Croix, 46MacDonald, W.D. 1980. Anomalous paleomagnetic di- U.S. Virgin Islands, 7th-11th August,1989, 472-480. rections in late Tertiary andesitic intrusions of the Miami Geological Society, Miami. Cauca depression, Colombian Andes. Tectonophysics, 34Harrington, H.J. 1962. Paleogeographic development of 68, 339-348. South America. American Association of Petroleum 47MacDonald, W.D., Doolan, B.L. & Cordani, U.G. 1971. Geologists Bulletin, 46, 1773-1814. Cretaceous-early Tertiary metamorphic K-Ar age val- 35James, K.H. 1990. The Venezuelan hydrocart>on habitat: ues from the south Caribbean. Geological Society of in Brooks, J. (ed.), Classic Petroleum Provinces. Geo- America Bulletin, 82, 1381-1388. logical Society Special Publication, 50, 9-35. 48MacDonald, W.D. & Hurley, P.M. 1969. Precambrian 36Jordan, T.H. 1975. The present-day motion of the Carib- gneisses from northern Colombia, South America. bean Plate. Journal of Geophysical Research, 80, Geological Society of America Bulletin, 80, 1867-1872. 4433-4439. 49MacDonald, W.D. & Opdyke, N.D. 1972. Tectonic rota- 37Julivert, M. 1970. Cover and basement tectonics in the tions suggested by paleomagnetic results from northern Cordillera Oriental of Colombia, South America, and a Colombia, South America Journal of Geophysical Re- comparison with some other folded chains. Geological search, 11,5720-5730.

245 STEPHEN K. DONOVAN

50MacDonald, W.D. & Opdyke, N.D. 1974. Triassic paleo- U-Pb and Rb-Sr systematics in the Imataca Series, magnetism of northern South America. American As- Guayana Shield, Venezuela. Earth and Planetary Sci- sociation of Petroleum Geologists Bulletin, 58, 208- ence Letters, 39, 281-290. 215. 62Morris, A.E.L., Taner, I., Meyerhoff, H.A. & Meyerhoff, 51MacDonald, W.D. & Opdyke, N.D. 1984. Preliminary A. A. 1990. Tectonic evolution of the Caribbean region; paleomagnetic results from Jurassic rocks of the Santa alternative hypothesis: in Dengo, G. & Case, I.E. (eds), Marta massif, Colombia. Geological Society of Amer- The Geology of North America. Volume H. The Carib- ica Memoir, 162, 295-298. bean Region, 433-457. Geological Society of America, 52Macellari, C. 1984. Late Tertiary tectonic history of the Boulder. 63 Tachira Depression, southwestern Venezuelan Andes. Muessig, K.W. 1978. The central Falcon igneous suite, Geological Society of America Memoir, 162, 333-341. Venezuela: alkaline basaltic intrusions of Oligocene- 53Mann, P., Schubert, C. & Burke, K. 1990. Review of Miocene age. Geologie en Mijnbouw, 57, 261-266. 64 Caribbean neotectonics: in Dengo, G. & Case, I.E. Muessig, K.W. 1984. Structure and Cenozoic tectonics of (eds), The Geology of North America. Volume H. The the Falcon Basin, Venezuela, and adjacent areas. Geo- Caribbean Region, 307-338. Geological Society of logical Society of America Memoir, 162, 217-230. 65 America, Boulder. Muessig, K.W. 1984. Paleomagnetic data on the basic 54Marriner, G.F. & Millward, D. 1984. The petrology and igneous intrusions of the central Falcon Basin, Vene- geochemistry of Cretaceous to Recent volcanism in zuela. Geological Society of America Memoir, 162, 231- Colombia: the magmatic history of an accretionary 237. plate margin. Journal of the Geological Society, Lon- 66Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of don, 141, 473-486. the Caribbean region; a plate-tectonic perspective: in 55Maze, W.B. 1984. Jurassic La Quinta Formation in the Dengo, G. & Case, I.E. (eds), The Geology of North Siena de Perija, northwestern Venezuela: geology and America. Volume H. The Caribbean Region, 405-432. tectonic environment of red beds and volcanic rocks. Geological Society of America, Boulder. 67 Geological Society of America Memoir, 162, 263-282. Poison, I.J. & Henao, D. 1968. The Santa Marta wrench 56Maze, W.B. & Margraves, R.B. 1984. Paleomagnetic re- fault: a rebuttal: in Saunders, J.B. (ed.), Transactions of sults from the Jurassic La Quinta Formation in the the Fourth Caribbean Geological Conference, Port-of- Perijarange, Venezuela, and their tectonic significance. Spain, Trinidad, 28th March-12th April 1965, 263-266. 68 Geological Society of America Memoir, 162, 287-293. Priem, H.N.A., Boelrijk, N. A.I.M., Hebeda, E.H., 57McClelland, L., Simkin, T., Summers, M., Nielsen, E. & Verdur-men, E.A.Th. & Verschure, R.H. 1973. Age of Stein, T.C. (eds). 1989. Global Volcanism 1975-1985. the Precambrian Roraima Formation in northeastern The First Decade of Reports from the Smithsonian South America: evidence from isotopic dating of Institution's Scientific Event Alert Network (SEAN). Roraima pyroclastic volcanic rocks in Suriname. Prentice Hall and American Geophysical Union, Geological Society of America Bulletin, 84, 1677-1684. Engle-wood Cliffs, New Jersey and Washington, D.C., 69Read, H.H. & Watson, J. 1975. Introduction to geology. vi+655 pp. Volume 2. Earth history. Part J: Early stages of Earth 58McCourt, W.J. & Aspden, J.A. 1987. A plate tectonic history. MacMillan, London, xii+221 pp. 70 model for the Phanerozoic evolution of central and Robertson, P. & Burke, K. 1989. Evolution of southern southern Colombia: in Duque-Caro, H. (ed.), Transac- Caribbean Plate Boundary, vicinity of Trinidad and tions of the Tenth Caribbean Geological Conference, Tobago. American Association of Petroleum Geologists Cartagena de Indias, Colombia, 14th-20th August, Bulletin, 73,490-509. 1983, 38-47. INGEOMINAS, Bogota 71Rod, E. 1956. Strike-slip faults of northern Venezuela. 59McCourt, W.J., Aspden, J.A. & Brook, M. 1984. New American Association of Petroleum Geologists Bulletin, geological and geochronological data from the Colom- 49, 451-416. 72 bian Andes: continental growth by multiple accretion. Schubert, C. 1982. Neotectonics of Bocono Fault, Journal of the Geological Society, London, 141, 831- western Venezuela. Tectonophysics, 85, 205-220. 73 845. Shagam, R. 1972. Andean research project, Venezuela: 60Montgomery, C.W. 1979. Uranium-lead geochronology principal data and tectonic implications. Geological of the Archean Imataca Series, Venezuelan Guayana Society of America Memoir, 132, 449-463. Shield. Contributions to Mineralogy and Petrology, 74Shagam, R. 1975. The northern termination of the Andes: 69, 167-176. in Nairn, A.E.M. & Stehli, F.G. (eds), The Ocean 61Montgomery, C.W. & Hurley, P.M. 1978. Total-rock

246 Northern South America

Basins and Margins. Volume 3. The Gulf of Mexico and & Cebula, G.T. 1974. Geologic evolution of the Sierra the Caribbean, 325-420. Plenum, New York. Nevada de Santa Marta, northeastern Colombia. Geo- 75Shagam, R., Kohn, B.P., Banks, P.O., Dasch, L.E., Var- logical Society of America Bulletin, 85, 273-284. gas, R., Rodriguez, G.I. & Pimentel, N. 1984. Tectonic 84Van Houten, F.B. 1976. Late Cenozoic volcaniclastic implications of Cretaceous-Pliocene fission-track ages deposits, Andean foredeep, Colombia. Geological So- from rocks of the circum-Maracaibo Basin region of ciety of America Bulletin, 87,481-495. Van Houten, western Venezuela and eastern Colombia Geological F.B. & James, H.E. 1984. Late Cenozoic Guayabo Society of America Memoir, 162,385-412. delta complex in southwestern Maracaibo Basin, 76Shepard, P.P. 1973. Sea floor off Magdalena Delta and northeastern Colombia. Geological Society of Santa Marta area, Colombia. Geological Society of America Memoir, 162, 325-332. America Bulletin,84, 1955-1972. 86Vasquez, E.E. & Dickey, P.A. 1972. Major faulting in 77Silver, E.A., Case, I.E. & MacGillavry, H.J. 1975. Geo- north-western Venezuela and its relation to global tec- physical study of the Venezuelan Borderland. Geologi- tonics: in Petzall, C. (ed.), Transactions of the Sixth cal Society of America Bulletin, 86, 213-226. Caribbean Geological Conference, Margarita, Vene- 78Skerlec, G.M. & Margraves, R.B. 1980. Tectonic signifi- zuela, 6th-14thJuly, 1971, 191-202. cance of paleomagnetic data from northern Venezuela. 87Vernette, G. & Klingebiel, A. 1988. Geometry and sedi- Journal of Geophysical Research, 85, 5303-5315. mentary facies of interdiapiric basins along the Carib- 79Steinitz, G. & Maze, W.B. 1984. K-Ar ages on horn- bean Colombian margin: in Barker, L. (ed.), blende-andesite from the Sierra de Perija, western Transactions of the Eleventh Caribbean Geological Venezuela. Geological Society of America Memoir, Conference, Dover Beach, Barbados, 20th-26th July, 162, 283-285. 1986,6:1-6:9. Energy and Natural Resources Division, 80 Sykes, L.R., McCann, W.R. & Kafka, A.L. 1982. Motion Barbados. of Caribbean Plate during last 7 million years and 88Viapiana, R.P. 1990. Eocene stratigraphical studies, Ma- implications for earlier Cenozoic movements. Journal racaibo Basin, northwestern Venezuela: in Larue, D.K. of'Geophysical Research, 87, 10656-10676. & Draper, G. (eds), Transactions of the Twelfth Carib- 81 Tarling, D.H. & Tarling, M.P. 1972. Continental drift: a bean Geological Conference, St. Croix, U.S. Virgin study of the Earth's moving surface. Pelican, Middle- Islands, 7th-llth August, 1989, 485-494. Miami sex, 142 pp. Geological Society, Miami. 82 Toussaint, J.F. & Restrepo, J.J. 1982. Magmatic evolution 89Wheeler, C.B. 1963. Oligocene and Lower Miocene stra- of the northwestern Andes of Colombia. Earth-Science tigraphy of western and northeastern Falcon Basin, Reviews, 18, 205-213. Venezuela. American Association of Petroleum Geolo- 83 Tschanz, C.M., Marvin, R.F., Cruz B., J., Mehnert, H.H. gists Bulletin, 47, 35-68.

247

CHAPTER 14

The Netherlands and Venezuelan Antilles

TREVOR A. JACKSON and EDWARD ROBINSON

Department of Geology, University of the West Indies, Mona, Kingston 7, Jamaica

INTRODUCTION bean Deformed Belt (SCDB) by the Los Roques Basin4,11.

THE SOUTHERN Caribbean, or Aruba-Blanquilla, chain includes the islands of the Netherlands and Venezuelan THE NETHERLANDS ANTILLES/ABC ISLANDS Antilles (Fig. 14.1). These islands are genetically related and lie across several wide platforms that have a northwest- Aruba southeast structural trend6. They are composed of Lower Aruba Lava Formation Cretaceous to Paleogene volcanic and sedimentary rocks The Cretaceous of Aruba consists of two major units. that were intruded by granitic plutons, and silicic and mafic The older is the Aruba Lava Formation, formerly called the dykes and sills. These rocks are overlain unconformably by Diabase-Schist-Tuff Foimation5, which is intruded by a Tertiary to Holocene sedimentary rocks. composite tonalitic pluton (Fig. 14.2a). The age of the Aruba A major gravity high parallels the island chain. This Lava Formation is based on Turonian(?) ammonites that regional high has been interpreted as a manifestation of were identified in sedimentary intercalations24. Rb-Sr iso- dense rock (that is, oceanic crust(?)), perhaps as much as 30 topic data for the tonalite body reveal an age that ranges 9,10,32,48,52 km thick . Individual highs characterise almost between 85.1 ± 0.5 Ma and 70.4 ±2.0 Ma40. However, the every island and are found in association with lows, which younger date is probably a reset age caused by a low-grade more or less coincide with the deep water passages between thermal event at about 72 Ma5,39. Additional isotopic data the islands. These features suggest that the islands are fault- for the pluton suggest that intrusion took place at 88±0.8 bounded, positive blocks, separated by sediment-filled pas- Ma39. sages and basins. The Aruba Lava Formation is over 3 km thick and The islands of the Netherlands Antilles, which com- consists of basalt, dolerite, volcaniclastic conglomerate, prise Aruba, Curacao and Bonaire, form part of the Leeward 11 sandstone, tuff, pelagic chert and cherty limestone. Basalt Antilles (ABC) terrane . Those of the Venezuelan Antilles, flows and dolerite sills form most of the formation. The further east, include the islands of Las Aves, Los Roques, basalts occur as pillowed lavas and sheet-flows, interdigi- La Orchila, La Blanquilla and Los Hermanos. Some of these tating with pyroclastic and volcaniclastic sediments5. The islands form part of the ABC terrane (Las Aves, Los Roques dolerite sills occur at various levels in the sequence and do and La Orchila), whereas those to the east (La Blanquilla 12 not attain a thickness greater than 300 m. and Los Hermanos) belong to the Blanquilla province . The basalt flows alternate with accretionary lapilli tuffs The islands of Margarita and Los Testigos, which belong to of basaltic composition and clasts of basalt and dolerite. A the Margarita subterrane, are not discussed in this chapter pebbly mudstone intercalated with fine-grained turbidites (see Donovan, Chapter 13, this volume). The ABC and contains ammonites of Turonian(?) age, together with a Blanquilla terranes are separated by the La Orchila Basin, shallow-water, benthic marine fauna which is a graben-like structure trending northwest-south- The basalts have been affected by contact metamor- east. The entire chain of islands is separated from mainland phism associated with the intrusion of the pluton, as there Venezuela by the Bonaire Basin and from the South Carib- are low to medium-grade metamorphic minerals such as

249 The Netherlands and Venezuelan Antilles

Figure 14.1. Generalized map of the southern Caribbean region. Key: A=Aruba; Bl=La Blanquilla; Bo=Bon- aire; C=Curacao; LA=Las Aves; LF=Los Frailes; LH=Los Hermanos; LM=Los Monjes; LO=La Orchila; LT=Los Testigos; M=Margarita; T=Tortuga; To=Tobago; Tr=Trinidad.

albite, epidote, hornblende and uralite present in the rocks. nopyroxene and magnetite. Trondhjemite occurs locally as However, despite the alteration of the basalt and dolerite, schlieren in the tonalite24. The schlieren range in length the original minerals in the mafics (clinopyroxene and pla- from 20 cm to more than 50 m and are commonly found in gioclase) are preserved. These minerals occur throughout close association with small pegmatite veins. Dykes of the sequence, but in different proportions. similar composition to the gabbros and trondhjemite intrude The major and trace element chemistry of the basalts the Aruba Lava Formation and the composite pluton. are similar in composition to the basalts in neighbouring The mineralogy of the trondhjemite comprises mainly Curacao. These mafic rocks, which show MORB affinities, plagioclase (An42-An22), together with quartz, and minor contain flat chrondrite-normalised REE patterns and differ potassium feldspar, biotite and hornblende. The pegmatite from their Curacao counterparts only by displaying a posi- contains micrographic intergrowths of quartz, plagioclase tive Eu anomaly5. feldspar, perthite and minor amounts of biotite. Tonalite Pluton From their geochemistry, the rocks that compose the The greater part of the island of Aruba is occupied by pluton show a calc-alkaline trend5. In addition to the behav- a pluton of batholithic proportions, composed primarily of iour of MgO and FeO with respect to SiO2, the REE patterns a hornblende tonalite, with lesser amounts of trondhjemite, are LREE-enriched with respect to the HREE. A mantle granitic pegmatite, norite and gabbro. The norite and gabbro source for the pluton was postulated by Priem et al.40 occur as roof pendants in the tonalite body and represent an because of the low initial 87Sr/86Sr ratios. earlier intrusive phase24. Their mineralogy grades from Both the Aruba Lava Formation and tonalite pluton equal amounts of hypersthene and augite in the norite to have been affected by late hydrothermal alteration. The abundant hornblende, relict augite and quartz in the gabbro. Aruba Lava Formation has also been metamorphosed by the Plagioclase, which is prominent in both rock types, is com- intrusion of the tonalite pluton and contains a zone of positionally zoned from An64 to An52, with cores sometimes alteration that ranges from hornblende-hornfels facies at the 5 being as calcic as An75. contact to prehnite-quartz facies some 4 km away . The tonalite is composed mainly of plagioclase crystals Limestones at Butucu and Seroe Colorado showing oscillatory zoning in the range An58-An37. Inter- A series of Tertiary limestones, with very restricted stitial quartz and potassium feldspar are the other felsic outcrop on Butucu Ranch, east-central Aruba, is distinct in minerals. The dark minerals are hornblende, biotite, cli- lacking quartz and containing a middle Tertiary faunal as-

250 T.A. JACKSON and E. ROBINSON

semblage50. The total exposed thickness is about 8.5 m and overlain conformably by the Cenozoic Midden-Curacao consists of pale brown to orange coralline algal-foraminif- Formation (Figs. 14.3b). The Curac ao Lava Formation con- eral packstone with minor coral fragments. The foraminifera tains a pelagic intercalation that yields six genera of ammon- include several species of Lepidocyclina, Heterostegina and ites of late middle Albian age51. K-Ar whole-rock dates for Pararotalia, collectively indicative of the Upper Eocene to the basalts in this formation seem anomalous, in that the possibly lowermost Oligocene (Fig.14.2b). Another iso- ages5 are either older or younger than Albian. lated exposure of what may be the same unit has been The Curacao Lava Formation includes over 5 km of reported from Seroe Colorado, at the southeastern tip of the pillow basalts, reworked hyaloclastites, dolerite sills, and a island, containing the coral Antiguastrea cellulosa (Upper thin succession of siliceous shales and limestones. The Eocene to Lower Miocene). formation is divided into lower, middle and upper sec- Seroe Domi Formation tions30, based mainly upon the petrographic and textural Named for exposures on Curacao, the Seroe Domi variations in the basalts. The lower part has a thickness of at Formation in Aruba is developed mainly along the south- least 1000 m in which picrite basalts alternate with tholei- western side of the island13 as a series of largely detrital itic basalts that display mostly variolitic texture. The middle limestones and conglomerates, to which can be added, as a 2000 m consists of tholeiitic pillow basalts which show member, the sands and clays of the Oranjestad borehole. predominantly non-variolitic texture, associated pillow Conventionally, the Seroe Domi Formation is divided into breccias and local dolerite sills. The upper part is compos ed three informal units, the older, middle and younger parts, almost entirely of non-variolitic pillow basalts, abundant but the full development of the formation is not seen on hyaloclastites, dolerite sills and thin bands of sedimentary Aruba. Here the older unit has been recorded only from rock. The picrites are composed principally of olivine and Seroe Canashito, where it consists of massive, extensively clinopyroxene30. Spinifex texture, commonly found in dolomitized, coral-algal limestone, lacking stratigraphically komatiites, has been identified in some of the picrites. diagnostic fossils13. Olivine and pyroxene, together with plagioclase and spinel, At Spaans Lagoen the Seroe Domi Formation is divided are the main minerals found in the variolitic and non-vari- into two parts13. An older part, of coral-poor, dolomitized olitic basalts. limestones, is massive at the base, but becomes thin-bedded The basalts of the Curacao Lava Formation are compo- in the overlying strata. It contains little terrigenous detritus, sitionally similar to the Aruba Lava Formation and are part of but has angular non-carbonate fragments in the basal layers. a Mid Ocean Ridge Basalt (MORB) sequence formed by The younger part is comprised of well-bedded limestones, shallow melting of mantle5. Based on the absence of ves- rich in terrigenous detritus, especially well-rounded peb- icles in the pillow lavas, the Curacao Lava Formation was bles, and with many macrofossils, particularly molluscs. extruded in a water depth of about 5 km30. The formation is Oranjestad Sands and Clays believed to be the ophiolitic analogue of layer 2, except that its The water borehole at Oranjestad was drilled to a depth thickness is much greater than that of normal oceanic crust, of 302 m in 1942, on the advice of two French friars, after aid is instead comparable with oceanic plateaus15. tests with a divining rod. Only hypersaline artesian water Knip Group was encountered, at 268 m. Microfossils from the borehole There is a marked angular unconformity that separates were analysed by Drooger17, who concluded that the sands the sedimentary rocks of the Knip Group from the underly- and clays were of Lower or Middle Miocene. However, a ing Curacao Lava Formation. The Knip Group consists review of Drooger's species list leads us to suggest that the almost entirely of pelagic, silica-rich and clastic sedimen- upper part of the borehole section is Pliocene age (based on tary rocks that contain a fauna of late Senonian age. Rocks of the presence of Pulleniatina obliquiloculata and Globoro- the Knip Group are well-dated biostratigraphically by talia miocenica) and that the lower part of the section (to rudists, larger and planktic foraminiferans, and ammonites. 268 m) is no lower than Upper Miocene (indicated by Ages range from Santonian or early Campanian for the basal Sphaeroidinella rutschi and Globogerinoides conglobatus). limestone lenses, through early to middle Campanian for the The sequence in the borehole probably represents a basal olistostrome units, to late Campanian and Maastrichtian for unit of the Seroe Domi Formation. the highest part of the sequence. Beets3 described two fossiliferous limestone lenses, the Zevenbergen and Cas- Curacao abao limestones, from the base of the Knip Group. The Curacao Lava Formation Zevenbergen limestone contains fragments of algae, corals, The Cretaceous stratigraphy of Curacao is represented gastropods, crinoids and rudists. The Casabao limestone is by two major units. The Curacao Lava Formation is uncon- composed of larger foraminifers, algae, bryozoans, crinoids formably overlain by the Knip Group, which in turn is and oysters. The presence of the rudist Torreites tschoppi in

251 The Netherlands and Venezuelan Antilles

Figure 14.2. (a) (top) Simplified geologic map of Aruba (modified from Helmers and Beets24), (b) (bottom) Stratigraphic column for Aruba.

252 T.A. JACKSON and E. ROBINSON

the Zevenbergen limestone suggests a late Santonian-Cam- race deposits of Quaternary age. panian age range for the limestone. Schaub43 named the Seroe di Cueba Formation for the The Knip Group ranges in thickness from just over well-known exposures forming the hill known as Cer'i 2000 m in the northwest to less than 100 m in the southeast Cueba (Fig.14.3a), previously called the -Seroe di Cueba of the island (Fig.14.3a). It is divided into nine formations, Series by Molengraaff36. Although not actually exposed, it is all of which are Upper Cretaceous3. In the northwest the Sint clear from the physiography that the limestones making up Christoffel Formation is a 1000 m thick succession of boul- the sequence form an outlier, unconformably overlying the der beds, slump breccias, pebbly mudstones, turbidites and rocks of the surrounding countryside, which belong to the silica-rich pelagic sediments, associated with limestone Knip Group. lenses. This formation grades upwards into Seroe Gratia Although the earliest palaeontological studies reported Formation, which is comprised of 300 m of cherry lime- ages ranging from Eocene to Oligocene, the Oligocene date stones, bedded cherts and radiolites, with minor amounts of for the upper part of the formation31 has never been con- volcaniclastic sandstone. The Seroe Gratia Formation is firmed. Many workers have attributed a late Eocene age to conformably overlain by the Lagoen Formation, which rep- the sequence, but this has also been questioned. For exam- resents a 750 m turbidite succession of sandstones, silt- ple, Molengraaff’s36 late Eocene date was based, at least stones, mudstones and marls with intercalations of tuffs. partly, on the presence of the echinoid Eupatagus grandi- Upwards and laterally these rocks grade into the Seroe florus (a junior synonym of E. clevei) in the lower limestone, Kortape and Zorgvlied Formations, which are both pelagic recorded previously from the limestones of St. silica-rich sequences. Towards the centre and southeast, Bartholomew in the northeastern Caribbean. Thought at that formations similar in lithology to those in the northwest, but time to be Upper Eocene, the St. Bartholomew limestones containing more clastic material, were described by Beets3. are now known to be Middle Eocene. The extensive larger These units are the Mameter Formation, with a maximum foraminiferal fauna from Cer'i Cueba closely resembles that thickness of 250 m in the centre, and the Stenen Koraal and of the Santa Rita Formation of western Venezuela, which Ronde Klip Formations in the southeast, which have a also was once thought to be late Eocene, but is now known maximum total thickness of 200 m. to be late middle Eocene in age28,38. Midden Curacao Formation The Mainsjie Formation is now poorly exposed, but Molengraaff 36 gave this name to a Cenozoic sequence consists of a series of calcareous sandy and conglomeratic over 1000 m in thickness, comprised of conglomerates, beds, at least 10 m thick, forming a hill in eastern Curacao. sandstones and shales, which outcrops in central and eastern The middle Eocene age of these beds was determined by Curacao. Beets2,3 described these beds as turbidites, with Hermes25 based on planktic foraminifera (Orbulinoides minor pelagic components, overlying the Knip Group con- beckmanni and/or Truncorotaloides rohri Biozones). formably. The vertical and lateral distribution of units Other exposures of Eocene rocks occur at Seroe Blanco, within the formation have been interpreted2,3 as indicating the mouth of a cave at Cueba Bosa, conglomeratic algal the existence of two interfingering submarine fans. A larger limestone at Plantation Patrick, and pebbly sandstones and fan bordering the northern flank of a basin (synclinorium of siltstones at Certu Kloof (Fig. 14.3a). All these are isolated Beets) contains unmetamorphosed and metamorphic from one another, but the available fossil evidence indicates (greenschist facies) basal andesite, together with rare, shal- that they can all be regarded as belonging to a single forma- low-water limestone components, derived from a northerly tion of probable late middle Eocene age, the same as for the source. The smaller fan, on the southern flank of the basin, Seroe di Cueba and Mainsjie Formations. The sedimentary contains the quartz-muscovite-plagioclase association seen rocks and fossils all indicate deposition under shallow main the underlying Knip Group, and was evidently derived rine conditions, ranging from locally brackish water and from a southerly source area. The early Paleocene (Danian) carbonate shoals to open shelf palaeoenvironments. age of the Midden Curacao Formation is based on poorly Seroe Domi Formation preserved planktic foraminifers (Globogerina? pseudobul- The Seroe Domi Formation represents an accumulation loides, G.? triloculinoides, Globorotalia? compressa). of reef and pebble debris as submarine talus fans. Molen- Seroe di Cueba and Mainsjie Formations graaff36 first named this unit for outcrops of hard to soft, The rocks considered here to be Eocene age consist of often coral-rich limestones, typically exposed at Seroe sequences, exposed at two localities, which have been given Domi, northwest of Willemstad. It is found in all three formational status, together with a number of other scattered islands of the Netherlands Antilles, mainly on the leeward exposures of various kinds. They have a common strati- (southwestern) sides, where it typically forms prominent graphic position, lying unconformably on older, usually cuestas, gently tilted towards the present coastline, and with Cretaceous rocks, and are overlain unconformably by ter- the landward cuesta margin often terminating in nearly

253 The Netherlands and Venezuelan Antilles

Figure 143. (a) (top) Simplified geologic map of Curacao (modified from Helmers and Beets24), (b) (bottom) Stratigraphic column for Curac ao.

254 T.A. JACKSON and E. ROBINSON

vertical cliffs. Intrusive Rocks De Buisonje13 restricted the term Seroe Domi Forma- There are two types of intrusions exposed on Curacao, tion to exclude limestones and other rocks forming late both of which are calc-alkaline in composition3. The oldest Quaternary constructional raised marine terraces. Because is a late Cretaceous diorite (K-Ar whole-rock age = 72 ± 7 the angle of dip of the strata forming the Seroe Domi Ma) which intrudes the Curacao Lava Formation. Younger Formation is normally greater than the angle of inclination intrusive rocks include a series of intermediate sills and of the cuesta tops, and where visible, the angle of inclination dykes that intrude the Knip Group and Midden Curacao of the formation base, the Seroe Domi Formation stratifica- Formation6. tion can be regarded as large scale foreset bedding. De These intrusives, together with their wallrocks, are Buisonje's evaluation of various sedimentary criteria, metamorphosed to zeolite facies. Laumontite, prehnite, mainly burrow infill data, also led him to conclude that there chlorite, pumpellyite, albite and calcite occur as replace- has been little or no tilting of the strata since deposition. The ment and vein minerals in these rocks. foreset stratification shows original depositional dip. For this reason de Buisonje used the informal terms older, Bonaire middle and younger, to describe the various parts of the Washikemba Formation formation, as the terms upper, middle and lower are difficult The Cretaceous units of Bonaire comprise the to apply. In general, in all three islands, he recognised the Washikemba and Rincon Formations (Fig. 14.4b). The for- older part, of strongly dolomitized algal limestones, poor in mer is over 5 km in thickness, consisting of pillow lavas, macrofossils; a middle part, rich in hermatypic corals and lava flows, shallow intrusions, and subaqueous pyroclastic algae, together with rounded, non-carbonate pebbles; and a flows that alternate with pelagic and volc aniclastic sedi- younger part, characterized by an abrupt increase in non-car- ments5. Imprints of ammonites identified by Cobban (in bonate pebbles. Beets and MacGillavry4) from cherty shales suggest a mid- In Curacao the older Seroe Domi Formation is reported dle late Albian age for this formation. The overlying Rincon only from the southeastern end of the island, at Tafelberg, Formation is late Senonian and comprises a 30 m succession Santa Barbara, Seroe Magasina, and Duivelsklip, where the of fossiliferous limestones and marls. basal layers locally contain larger foraminiferans, such as The sedimentary rocks that make up the Washikemba Miogypsina globulina (Drooger in de Buisonje13). At Tafel- Formation appear to have accumulated in a deep water berg (Fig. 14.3a) the section has undergone locally extensive environment5. Pelagic sediments in the lower half are cherts, phosphatization23. radiolarites and cherty shales, whereas cherty limestones The middle and younger parts of the formation are best and marls are common in the upper half. Subaqueous pyro- considered collectively as they occur together at several clastic flows, consisting largely of angular pumice frag- localities west of Tafelberg in the absence of the older part ments, commonly occur in the upper part of the formation. of the formation. Coral-algal bioclastic limestones, includ- Although there is a broad range in the composition of ing corals such as Montastrea, pass up into layers with the volcanic rocks from basalt to rhyolite, their overall benthic molluscs, such as Ostrea and Spondylus. The high- volume is bimodal, with basaltic andesite and rhyolite being est parts of the middle sequence contain Acropora, some- the dominant types. In addition to the pyroclastic rocks, the times in abundance. The younger part of the Seroe Domi volcanic rocks occur as pillow lava flows, sheet-flows, Formation consists of thick conglomerates with well- laccoliths and sills. The chemistry of these rocks indicate an rounded pebbles. island-arc derivation for the sequence. Donnelly and Ro- On the basis of the Miogypsina found at Tafelberg a gers16 and Donnelly et al.15 described the volcanic rocks as late-early Miocene to early-middle Miocene age can be being part of their Primitive Island Arc (PIA) series. These suggested for the oldest part of the Seroe Domi Formation. rocks are therefore different in their tectonomagmatic set- However, mixed basal faunas have been collected at other ting to the volcanic rocks of the Curacao Lava and Aruba localities, so it is possible that the larger foraminiferans are Lava Formations. all reworked from older formations. The age of the middle The rocks of the Washikemba Formation have been Seroe Domi Formation is limited from late Miocene (or affected by low-grade metamorphism. Primary minerals in later) to late Pliocene, on the evidence of the coral faunas the volcanic rocks, such as pyroxene, hornblende and pla- (Agaricia, Stylophora, Mussa angulosa, possibly Mean- gioclase, have been altered to low-grade mineral assem- drina meandrites18); the presence ofAcropora cervicornis bleges of the prehnite-pumpellyite and zeolite facies4. in the younger Seroe Domi Formation strongly suggests an Rincon Formation early Pleistocene age. The Rincon Formation occurs as isolated outcrops in the central part of the island, and is comprised of limestone,

255 The Netherlands and Venezuelan Antilles

Figure 14.4. (a) (top) Simplified geologic map of Bonaire (modified from Helmers and Beets24), (b) (bottom) Stratigraphic column for Bonaire. 256 T.A. JACKSON and E. ROBINSON

marl and calcareous sandstone. Its stratigraphic relationship foraminiferal assemblages from the Porta Spanjo borehole with the underlying Washikemba Formation is not clear due suggest an outer sublittoral to bathyal palaeoenvironment25. to the faulted nature of the contacts. However, Beets and Seroe Domi Formation MacGillavry4 inferred that the Rincon Formation lies un- The Seroe Domi Formation is divided into three parts conformably on the Washikemba Formation, because of as on Curacao and Aruba13. Original dips of more than 30° their different ages, as determined from their fossil assem- have been recorded. The older part is not as dolomitic as on blages. The shallow-water fauna, which includes rudists and the other islands and in places contains a basal, angular, other bivalves, algae, gastropods and larger foraminiferans, volcanic breccia, including blocks with mid Tertiary larger indicate a middle to late Maastrichtian age. foraminifers. As in the other islands the fossil evidence Soebi Blanco Formation and Pos Dominica indicates a late Miocene to early Pleistocene age for the unit. conglomerates The early Miocene fossils sometimes encountered in the The Soebi Blanco Formation consists of 400 m of base of the unit are either derived from an early Miocene conglomerates, sandstones and mudstones at Seroe Largo in unit, no longer exposed on the islands, or may be compo- central Bonaire (Fig.14.4a). Fragments include quartzites, nents of a local early Miocene unit, unconformable below gneisses and schists, material which is exotic to Bonaire, but the Seroe Domi Formation. As in Curacao, the Seroe Domi resembles some of the components found in the Midden of Bonaire is indicative of deposition as talus in a deep Curacao Formation. U-Pb analysis on zircon fractions from forereef palaeoenvironment. a granulite pebble contained in the formation revealed a date Quaternary of the Netherlands Antilles of approximately 1150 Ma41, which would suggest that the A sequence of denudational terraces, or benches, is source rock originated on mainland South America. Creta- widely recognised throughout the islands of Aruba, Curacao ceous limestone fragments also occur. This sequence has and Bonaire13,26. Many of these are associated with con- generally been regarded as fluviatile in origin. Red constructional terrace limestones. These evidently formed by in glomerates with granodiorite pebbles at Pos Dominica are situ deposition of coral reef complexes fringing a slowly also of fluvial origin and may be related. However, these uplifting landscape, while experiencing a series of late Pleis- rocks lack the metamorphic fragments of the Soebi Blanco tocene eustatic sea-level fluctuations. Formation, and pass upwards into mollusc-bearing, calcare- De Buisonje13 recognised five terrace levels, named, ous sandstones. from highest to lowest; Highest Terrace, Higher Terrace, The unfossiliferous Soebi Blanco Formation is usually Middle Terrace II, Middle Terrace I, and Lower Terrace. regarded as being of Paleocene age (Fig. 14.4b) on the basis Herweijer and Focke26 distinguished more than ten of its stratigraphic position and because it contains similar benches/terraces associated with these terrace limestones. lithic components to the Paleocene Midden Curacao Forma- They showed that both the Lower Terrace and Middle tion. However, the depositional environment is apparently Terrace I are composite terrace systems, made up of several quite different and a direct palaeogeographic connection distinct limestone units separated by unconformities (five in appears unlikely because a southern origin is indicated for the Lower Terrace and at least four in Middle Terrace I). the Midden Curacao Formation. The Pos Dominica red beds In the Lower Terrace limestones the best preserved are probably of Eocene age, based on the molluscan fauna deposits are divided into two zones by Herweijer and of the overlying beds29. Focke26. A barrier -reef zone, made up of almost exclusively of Montagne Formation a framework dominated by the coral Acropora palmata and These rocks are a series of impure, yellow -weathering the alga Porolithon pachydermum, is distinguished from a limestones that outcrop southwest of Montagne. Similar back-reef zone, dominated byMontastrea annularis and exposures have been reported from east of the Rincon de- other corals, such as Siderastrea, Diploria and Acropora pression, from a borehole at Porta Spanjo and from overly- cervicornis, often forming an interlocking framework, ing the Soebi Blanco beds at Seroe Largo33. At Montagne associated with fragmental, coral-algal sediments. Thick (Fig. 14.4a) the formation consists of a lower, mollusc-bear- coralline algal crusts and massive rhodolites indicate open ing unit, overlain by an upper unit characterized by echi- marine, relatively high energy regimes. noids and larger foraminifers. The assemblages are The elevation of the five limestone units of the Lower indicative of a late to middle Eocene age, and an inner to Terrace range from +10 m to -9 m a.s.1. (above sea level), middle sublittoral palaeoenvironment. In contrast, a small while those of the Middle Terrace range up to 27 m a.s.1.. exposure nearby (seen by ER) includes planktic foraminif- The higher terraces are more fragmentary and less well-pre- eral, sponge spiculites of a decidedly bathyal character and served. of late Eocene age. Structural complications prevent a clear Uranium series dates on the Lower Terrace group range interpretation of the stratigraphic relationships. The planktic from 90,000 to 13 0,000 years bp, indicating correlation with

257 The Netherlands and Venezuelan Antilles

similar aged deposits on La Orchila and La Blanquilla of the La Blanquilla on the Los Hermanos islands. These rocks, Venezuelan Antilles, and with the globally recognised which are believed to be the wallrock for the Garanton 125,000 years bp sea-level high, generally correlated with batholith of La Blanquilla, reveal anomalous K-Ar dates of the Sangamon Interglacial Stage. Uranium and thorium between 67 and 71 Ma42. Pegmatite dykes that cross-cut the series ages determined for the Middle Terrace I group (up metamorphic rocks may be responsible for the low ages to 27 m above sealevel) are less conclusive, but suggest an obtained in the metamorphic rocks. age of about 500,000 years bp. Quaternary More recent events include the formation of Caracas- On Gran Roque there are conglomerate terraces of 3 and 5 baai in southeastern Curac ao, apparently the scar left by a m elevation, consisting of fragments from the underlying large coastal landslip, involving about 0.25 km of the Seroe basement rocks, cemented by carbonate45. On La Orchila Domi deposits, which slid into the adjacent ocean to depths erosional features, indicating three old sea-level elevations, of up to 800 m. The age of this event has been placed at 5,000 were identified by Schubert and Valastro47 at 9, 11.5 and 33 to 10,000 years bp14. m above present sea level. Some of these are associated with marine or coastal deposits. The most extensive level is THE VENEZUELAN ANTILLES associated with various deposits including storm debris near the coast. Corals and the gastropod Strombus are prominent Many of the islands that constitute the Venezuelan Antilles fossil components. Two230Th/238U dates on the corals from are atolls with little exposed area above sea level, such as this terrace yielded ages of about 131,000 years bp. Las Aves. Only on the islands of Gran Roque, La Orchila, By far the most extensive series of marine terraces is La Blanquilla, Los Hermanos and Los Testigos are there exposed on the island of La Blanquilla. They are at three exposures of Cretaceous and younger rock (Fig. 14.5). On levels, 7-10, 11-15 and 25 m above sea level. The mainly all of these islands Cretaceous igneous and metamorphic carbonate deposits, known collectively as the La Blanquilla rocks are unconformably overlain by Quaternary sediments Formation, rest unconformably on the Garanton (Fig. 14.5). Tertiaiy rocks have not been found. trondhjemite20,44. The highest terrace surface has developed doline topography on the extensively recrystallised lime- Cretaceous stones of which it is formed. A basal conglomerate includes According to Schubert and Moticska45, the oldest rocks in blocks of amphibolite, not found elsewhere on the island, the Venezuelan Antilles outcrop on La Orchila in the form of but exposed as basement in the Los Hermanos archipelago, chlorite schists and phyllites, quartz-epidote-garnet or- 15 km to the southeast. The middle terrace covers about a thoamphibolite, hornblende gneiss, and micaceous epidote third of the island and exhibits two facies; a coral-rich facies gneiss and schist. These rocks are intruded by a composite considered a reef, and a sandy facies interpreted as lagoonal pluton, ranging from granite to dolerite, which has produced a and beach deposits behind the reef. Schubert and Szabo46 low-grade contact metamorphic aureole in the surrounding obtaineda 234U/238Uage of 325,000 ±70,000 years bp from wallrocks. Ultramafic rocks such as peridotite and ser- a Diploria coral. pentinite are exposed in the centre of the island, but their The youngest (lowest) terrace, also known as the El 34 relationship with the other rocks is unclear. Falucho Member of the La Blanquilla Formation , out- Wes t of La Orchila, on the island of Gran Roque, mafic crops around most of the island. The deposits consist almost igneous rocks have been intruded by quartz-diorite dykes. entirely of coral fragments, the rocks showing much less These silicic rocks, which give K-Ar dates of between 66 diagenesis than those of the other terraces. Schubert and 46 230 and 77 ± 6 Ma6, are intruded by pegmatite and aplite dykes. Szabo obtained a Th age of 133,000 ± 7,000 years bp Hargraves and Skerlac 22 identified hornblende, biotite, minor from a Montastrea coral, indicating correlation of this set of clinopyroxene, slightly altered feldspar, and quartz as the terrace deposits with those of the lowest terrace on La major rock-forming minerals in the quartz-diorite. Orchila Schubert and Moticska45 reported that the rocks on La Blanquilla are composed predominantly of a trondhjemite- GEOLOGICAL EVOLUTION OF THE SOUTH tonalite pluton called the Garanton batholith. Bellizzia and CARIBBEAN ISLAND CHAIN Dengo6 cited K-Ar dates of between 81 and 64 Ma for these rocks, which were in turn intruded by pegmatite dykes and The paucity of good geochronometric control, especially for veins. Schubert and Moticska45 described the occurrence of some of the rock types of the Venezuelan Antilles, places amphibolite xenoliths within the tonalitic part of the pluton. constraints on reconstructing the evolution of the South Outcrops of amphibolite, together with hornblende Caribbean island chain during the Mesozoic era. If the gneiss and biotite-epidote schist, occur 15 km southeast of chlorite schists, hornblende gneisses, and amphibolites of

258 T.A. JACKSON and E. ROBINSON

Figure 14.5. Geologic maps of Gran Roque, La Blanquilla, La Orchila and Los Hermanos (after Santamaria and Schubert42).

259 The Netherlands and Venezuelan Antilles

260 T.A. JACKSON and E. ROBINSON

La Orchila and Los Hermanos are correlated with the North direction for the fluviatile Soebi Blanco Formation has not Coast Schist Group of Tobago, further east (see Jackson and been established. U-Pb ages of 1150 Ma obtained by Priem Donovan, Chapter 11, this volume), then these rocks repre- et al.41 for fragments in the Soebi Blanco Formation have sent the oldest exposed rocks along the island chain. This been used by those authors to suggest a possible source in interpretation is based on similarities in lithology and meta- the present Guajira Peninsula, where rocks of this age and morphic grade, but ignores the K-Ar radiometric ages for character are exposed today. According to Maresch and 42 the hornblende gneiss of Los Hermanos, which has been Santamaria and Schubert , during the Danian to middle dated at between 67 and 71 Ma42. We suspect that these are Eocene, rocks in this region of the chain were subjected to reset dates caused by the intrusion of the trondhjemite- folding, low-grade regional metamorphism, and intrusion tonalite pluton in the late Cretaceous. These rocks therefore by minor dykes and sills. represent part of the first phase of arc magmatism that The early and middle Eocene was a time of high relative 21 occurred in the late Jurassic(?)-early Cretaceous49. They eustatic sea-level , and deposits of this age are common were subsequently deformed and metamorphosed prior to a throughout northern South America. In the offshore region second phase of magmatism in the mid-Cretaceous. of the Netherlands Antilles, scattered records of middle The mafic rocks of Araba, Curacao and Gran Roque, Eocene shelf carbonate and non-carbonate deposition show which are described as ocean floor tholeiites/MORB, repre- these islands to have been positive features compared with sent a mid-Cretaceous magmatic phase. According to San- the deeper water palaeoenvironments displayed by most tamaria and Schubert42 their presence indicates that there similar-aged units in northern Venezuela (Pauji/Mene 19,27 was an underthrusting of the Caribbean plate beneath South Grande, Guacharaca beds ), and the inshore island of 7,28 America at this time. However, more recent work5,49 sug- Margarita (Punta Mosquito Formation ). In the Venezuelan gests that these rocks originated as a result of back-arc/in- offshore islands this depositional phase is not represented. tra-arc spreading, whic h Donnelly et al.15 regard as a Records for the middle Tertiary are also mostly lacking, as manifestation of oceanic plateau magmatism. Accompany- they are over large parts of northern South America, except in the centres of the newly forming pull-apart basins ing ocean floor volcanism was the eruption of PIA volcanic 1,8,37 rocks as indicated by the Washikemba Formation of Bon- of Falcon and in the offshore areas . aire. These volcanic rocks are bimodal in composition and Records for the late Tertiary to early Quaternary, typi- are comprised of basaltic andesite, rhyolite lava flows and fied by the Seroe Domi Formation of the ABC islands, submarine pyroclastic deposits, thereby indicating a domi- suggest continuing positive areas with fringing reefs devel- nance of subaqueous arc activity at this time. Together with oping from late Miocene times onwards and bordering the the MORB of Curacao, Aruba, and Gran Roque, these rocks Neogene pull-apart basins. form the remnant of a mid-Cretaceous arc terrane. Compared with data reported from Barbados, the aver- The youngest phase of magmatism in the Mesozoic is age rate of uplift of the three Netherlands Antilles islands well represented throughout the Southern Caribbean island over the last half million years has been very low, about 0.05 m/1000 yr, compared with about 0.43 m/1000 yr for chain in the form of late Cretaceous -early Tertiary granitoid 26 batholiths of calc -alkaline composition5,42,49. The plutons Barbados . display no major geochemical variations along the island chain; however, when compared with the Mesozoic plutons ACKNOWLEDGEMENTS—The authors thank Carlos Schubert for allowing of the Caribbean Mountains of Venezuela there is a marked us to reproduce Fig. 14.5. We are also indebted to Steve Donovan and two anonymous referees for their critical comments of the manuscript. potassium polarity between the two regions42 This geochemical polarity prompted Santamaria and Schubert to conclude that during the late Cretaceous the South Carib- REFERENCES bean islands continued to be the site of island-arc magmatism, whereas the high-potassium plutons of the Caribbean 1Amstein, R. 1987. Transgresion Miocena en la region Mountains were associated with continental arc magma- norcentral de Falcon. Memoria de VII Congreso tism. Geologico Venezolano, 2, 641-661. By the Paleocene igneous activity was on the wane, and 2Beets, D.J. 1972. Lithology and stratigraphy of the Creta- the region of Curacao and Bonaire was receiving sediments ceous and Danian succession of Curacao. Uitgaven from a continental source. In Curacao these sedimentary Natuurwetenschappelijke Studierkring voor Suriname rocks, represented by the Midden Curacao Formation, were enNederlandseAntillen, Utrecht, 70,153 pp. being deposited from a southern source into the southern 3Beets, D.J. 1977. Cretaceous and early Tertiary of Curacao: part of a basin, while simultaneously being fed with island- in Eighth Caribbean Geological Conference Guide to arc debris from a northern source. In Bonaire the source the Field Excursions on Curacao, Bonaire and Aruba. 261 The Netherlands and Venezuelan Antilles

GUA Papers of Geology, 10, 7-17. tectonic setting of Caribbean magmatism: in Dengo, G. 4Beets, D J. & MacGillavry, HJ. 1977. Outline of the Cre- & Case, J.E. (eds), The Geology of North America. taceous and early Tertiary history of Curacao, Bonaire Volume H. The Caribbean Region, 339-374. Geologi- and Aruba: in Eighth Caribbean Geological Confer- cal Society of America, Boulder. ence Guide to the Field Excursions on Curacao, Bon- 16Donnelly, T.W. & Rogers, J.J.W. 1978. The distribution aire and Aruba. GUA Papers of Geology, 10,1-6. of igneous rock suites throughout the Caribbean. 5Beets, D.J., Maresch, W.V., Klaver, G.T., Mottana, A., Geologie en Mijnbouw, 57, 151-162. Bocchio, R., Beunk, F.F. & Monen, H.P. 1984. Mag- 17Drooger, C.W. 1953. Miocene and Pleistocene foraminif- netic rock series and high pressure metamorphism as era from Oranjestad, Aruba (Netherlands Antilles). constraints on the tectonic history of the southern Car - Contributions to the Cushman Foundation for ibbean. Geological Society of America Memoir, 162, Foraminiferal Research, 4 (4), 116-147. 95-130. 18Frost, S.H. 1972. Evolution of Cenozoic Caribbean coral 6Bellizzia, A. & Dengo, G. 1990. The Caribbean mountain faunas: in Petzall, C. (ed), Transactions of the Sixth system, northern South America; a summary: in Dengo, Caribbean Geological Conference, Margarita, Vene- G. & Case, J.E. (eds), The Geology of North America. zuela, 6th-14thJuly, 1971, 461-464. Volume H. The Caribbean Region, 167-176. Geological 19Furrer, M. 1968. The depositional environment of the Society of America, Boulder. Mene Grande Formation. Asociacion Venezolano de 7Bermudez, P.J. & Gamez, H.A. 1966. Estudio paleon- Geologia, Mineriay Petroleo, Boletinlnformativo, 10, tologico de una section del Eoceno. Memoria de la 192-195. Sociedad de Ciencias Naturales La Salle, 26 (75), 205- 20Gonzalez de Juana, C., Iturralde de Arozena, J.M., & 259. Picard Cadilat, X. 1980. Geologia de Venezuela yde 8 Boesi, T. & Goddard, D. 1990. A new geologic model sus Cuencas Petroliferos (2 volumes). Ediclones Fon- related to the distribution of hydrocarbon source rocks in inves, Caracas. the Falcon Basin, northwestern Venezuela: in Biddle, 21Haq, B.U., Hardenbol, J. & Vail, P.R. 1987. Chronology K.T. (ed.), Active Margin Basins. American Associa- of fluctuating sea levels since the Triassic. Science, 235, tion of Petroleum Geologists, Memoir, 52, 303-319. 1156-1167. 9 Bonini, W.E., Pimstein de Gaeta, C. & Graterol, V. 1977. 22Hargraves, R.B. & Skerlec, G.M. 1982. Paleomagnetism Mapa de anomalias gravimetricas de Bouguer de la of some Cretaceous -Tertiary igneous rocks on Vene- parte norte de Venezuela y areas vacinas, Escala zuelan offshore islands, Netherlands Antilles, Trinidad 1:100,000. Ministerio de Energia y Minas, Caracas. and Tobago: in Snow, W., Gil, N., Llinas, R., Ro- 10 Case, J.E. 1975. Geophysical studies in the Caribbean driguez-Torres, R., Seaward, M. & Tavares, I. (eds), Sea: in Nairn, A.E.M. & Stehli, F.G. (eds), The Ocean Transactions of the Ninth Caribbean Geological Con- Basins and Margins. 3. The Gulf of Mexico and the ference, Santo Domingo, Dominican Republic, 16th- Caribbean. Plenum, New York, 107-181. 20th August, 1980 ,2, 509-518. H Case, J.E., Holcombe, T.L. & Martin, R.G. 1984. Map of 23Have, T. ten, Heijnen, W. & Nickel, E. 1982. Alterations geologic provinces in the Caribbean region. Geological in guano phosphates and Mio-Pliocene carbonates of Society of America Memoir, 162, 1-30. Table Mountain Santa Barbara, Curacao. Sedimentary 12 Case, J.E., MacDonald, W.D. & Fox, P.J. 1990. Caribbean Geology, 31, 141-165. crustal provinces; seismic and gravity evidence: in 24Helmers, H. & Beets, D.J. 1977. Cretaceous of Aruba: in Dengo, G. & Case, J.E. (eds), The Geology of North Eighth Caribbean Geological Conference Guide to America. Volume H. The Caribbean Region, 15-36. the Field Excursions on Curacao, Bonaire and Geological Society of America, Boulder. Aruba. GUA Papers of Geology, 10, 29-35. 13 de Buisonje, P.H. 1974. Neogene and Quaternary geology of 25Hermes, J.J. 1968. Planktonic foraminifera from the Seroe Aruba, Curacao and Bonaire (Netherlands Antilles). Mainsji Formation of Curacao. Geologie en Mijnbouw, Uitgaven Natuurwetenschappelijke Studiekring voor 47,280-290. Suriname en de Nederlandse Antillen, 78, 291 pp. 26 14 Herweijer, J.P. & Focke, J.W. 1978. Late Pleistocene de Buisonje, P.H. & Zonneveld, J.I.S. 1976. Caracasbaai: a depositional and denudation history of Aruba, Bonaire submarine slide of a high coastal fragment in Curacao. and Curac ao. Geologie en Mijnbouw, 57, 177-187. Nieuwe West-Indischegids, 5, 55-88. 27Hunter, V.F. 1972. A middle Eocene flysh from east 15 Donnelly, T.W., Beets, D., Carr, M.J., Jackson, T., Klaver, G., Falcon, Venezuela; in Petzall, C. (ed.), Transactions Lewis, J., Mauiy, R., Schellenkens, H., Smith, A.L., of the Sixth Caribbean Geological Conference, Wadge, G, & Westercamp, D. 1990. History and Margarita, Venezuela, 6th-14thJuly, 1971, 126-130. 262 T.A. JACKSON and E. ROBINSON

28Hunter, V.F. 1978. Foraminiferal correlation of Tertiary tilles. Geologie en Mijnbouw, 57,293-296. mollusc horizons of the southern Caribbean area. 41Priem, H.N.A, Beets, D.J. & Verdurmen, E.A.T. 1986. Geologie en Mijnbouw, 57,193-203. Precambrian rocks in an early Tertiary conglomerate on 29Jung, P. 1974. Eocene mollusks from Curacao, West Bonaire, Netherlands Antilles (southern Caribbean bor- Indies. Verhandlungen der Naturforschenden Gesell- derland): evidence for a 300 km eastward displacement schqftin Basel, 84, 483-500. relative to the South American mainland? Geologie en 30Klaver, G.T. 1987. The Curacao Lava Formation: an Mijnbouw, 65,35-40. ophiolitic analogue of the anomalous thick layer 2B 42Santamaria, F. & Schubert, C. 1974. Geochemistry and of the mid-Cretaceous oceanic plateaus in the geochronology of the southern Caribbean-northern western Pacific and central Caribbean. GUA Papers of Venezuala plate boundary. Geological Society of Amer- Geology, 27,168pp. ica Bulletin, 85, 1085-1098. 31Koch, R. 1929. Berichtigung und Erganzung zu der Notiz 43Schaub, H.P. 1948. Geological observations on Curac ao, "Tertiarer Foraminiferankalk von der Insel Curacao". N.W.I. American Association ofPetroleum Geologists, Eclogae Geologicae Helvetiae, 22, 159-161. 32, 1275-1291. 32Lagaay, R.A. 1969. Geophysical investigations of the 44Schubert, C. 1976. Formacion Blanquilla, isla La Blan- Netherlands Leeward Antilles. Verhandelingen der quilla (Dependencias federates). Informe preliminar Koninklijke Nederlandse Akademie van Wetenschap- sobre terrazas cuaternarias. Acta Cientifica Venezo- pen, Afd. Natuurkunde, 25 (2), 86 pp. lana, 27, 251-257. 33MacGillavry, HJ. 1977. Tertiary Formations: in Eighth 45Schubert, C. & Moticska, P. 1972. Geological recon- Caribbean Geological Conference Guide to the Field naisance of the Venezuelan islands the Caribbean Sea Excursions on Curacao, Bonaire and Aruba. GUA between Los Roques and Los Testigos: in Petzall, C Papers of Geology, 10, 36-38. (ed.), Transactions of the Sixth Caribbean Geological 34Maloney, N.J. 1971. Geologia de la isla de la Blanquilla y Conference, Margarita, Venezuela, 6th-14th July, notes sobre el archipelago de Los Hermanos, Venezuela 1971, 81-82. oriental. Ada Cientiflca Venezolana, 22, 6-10. 46Schubert, C. & Szabo, BJ. 1978. Uranium-series ages 35Maresch, W.V. 1974. Plate tectonics origin of the Carib- of Pleistocene marine deposits on the islands of bean mountain system of northern South America: dis - Curacao and La Blanquilla, Caribbean Sea. cussion and proposal. Geological Society of America Geologie en Mijnbouw, 57, 325-332. Bulletin, 85, 669-682. 47Schubert, C. & Valastro, S. 1976. Quaternary geology of 36 Molengraaf,G.J.H. 1929. Geologie en geohydrologie van La Orchila island, central Venezuelan offshore, Carib- het eiland Curacao. Waltman, Delft, 126 pp. bean Sea. Geological Society of America Bulletin, 87, 37Muessig, K.W. 1984. Structure and Cenozoic tectonics of 1131-1142. the Falc6n Basin, Venezuela, and adjacent areas. Geo- 48Silver, E.A., Case, I.E. & MacGillavry, HJ. 1975. Geo- logical Society of America Memoir, 162, 217-230. physical study of the Venezuelan borderland. Geologi- 38 Pittelli Viapiana, R. 1990. Eocene stratigraphical studies, cal Society of America Bulletin, 86, 213-226. Maracaibo Basin, Northwestern Venezuela: in Larue, 49Snoke, A.W. 1991. An evaluation of the petrogenesis of D.K. & Draper, G. (eds), Transactions of the Twelfth the accreted Mesozoic island arc of the southern Carib- Caribbean Geological Conference, StCroix, U.S. Virgin bean: in Gellizeau, K.A. (ed.), Transactions of the Islands, 7th-llth August, 1989, 485-494. Miami Second Geological Conference of the Geological Soci- Geological Society, Florida. ety of Trinidad and Tobago, Port-of-Spain, 3rd-8th 39 Priem, H.N.A., Beets, D J., Boelrijk, N.A.I.M. & Hebeda, April 1990, 222-232. E.H. 1986. On the age of the late Cretaceous tonali- 50Westermann, J.H. 1932. The geology of Aruba. Unpub- tic/gabbroic batholith on Aruba, Netherlands Antilles lished Ph.D. thesis, Utrecht University, 129 pp. (southern Caribbean borderland). Geologie en 51Wiedmann, J. 1978. Ammonites from the Curacao Lava Mynbouw, 65,247-256. Formation, Curacao, Caribbean. Geologie en 40 Priem, H.N.A., Beets, D.J., Boelrijk, N.A.I.M., Hebeda, Mijnbouw, 57,361-364. E.H., Verfuimen, E.A.T. & Veischure, R.H. 1978. 52Worzel, J.L. 1965. Pendulum gravity measurements at Rb-Sr evidence for episodic intrusion of the late Creta- sea, 1936-1959. John Wiley & Sons, New York, ceous tonalitic batholith of Aruba, Netherlands An- 422pp.

263

CHAPTER 15

Northern Central America

BURKE BURKART

Department of Geology, University of Texas at Arlington, Arlington, Texas 76019, U.SA.

INTRODUCTION nisms, presenting the first model of interactions between these plates, and their model is still useful in describing plate A GOAL of this chapter is to present a current picture of the interactions involving Central America. Recent studies have geology of northern Central America, a region which en- adopted the designation proposed by Dengo27,28 that di- compasses a large part of the western margin of the Carib- vides Central America into the Maya block (or Yucatan bean and shares its complex geological history. Another block) north of the major transform faults, and the Chortis purpose is to discuss constraints Central American geology block to the south (Fig. 15.2). may place on plate tectonic models, many of which are Hess and Maxwell47 proposed a left-lateral displace- discussed in other chapters. ment of northern Central America of about 600 km across Northern Central America is defined for this chapter as faults extending westward from the Cayman Trough. Since the region from the Isthmus of Tehuantepec to Nicaragua, that time, data from the Caribbean generally have supported including the Yucatan Peninsula of Mexico, all of Guate- models of eastward displacement of the Caribbean Plate past mala, Belize, and Honduras (Fig. 15.1). The Peten region of northernmost Central America, while geological mapping Guatemala is the northernmost part or 'panhandle' of that in Guatemala and Chiapas, Mexico has verified a set of country, located just to the west of Belize. In common usage left-lateral strike-slip faults that appear to be prolongations the term northern Guatemala refers to the region adjoining of boundary faults of the Cayman Trough. Yet, if the ulti- the Peten to the south. mate test of a model of displaced terranes is the strong Few geological studies preceded the monumental matching of lithology, stratigraphy, or structure across a works of Karl Sapper, which span the interval from 1890 to plate boundary zone, then the model of Hess and Maxwell47, 1937, and provided the foundation for the geology of Cen- and other more recent models with even greater magnitudes tral America, Belize and southern Mexico84. Sapper's 1937 85 of offset, remain unproven. publication included the first published geologic map of The Maya block is tacitly regarded as an independent Central America. Roberts and Irving provided an excel- subdivision of the North American Plate, separated on the lent summary of the geology of Central America in their northwest by the poorly-defined, north-trending, left-lateral treatise on the mineral deposits of the region. There are two Salina Cruz fault89, by normal faults separating it from the excellent, comprehensive works on the geology of Central 95 34 Yucatan Basin, and a generally unspecified set of structures America, by Weyl and by Donnelly et al. . from the Gulf of Mexico Basin82. In more recent studies, Burkart and Scotese12 have mapped a complex zone of PLATE TECTONIC SETTING right-lateral en echelon faults they call the Orizaba fault zone. These faults run northwest along the western margin Northern Central America is generally accepted to be the of the Isthmus of Tehuantepec, forming the probable north- place of intersection of the Cocos, North American, and western boundary of the Maya block (Fig. 15.3). The north- Caribbean plates44. The intersection is a complex arrange- ern margin along coastal and offshore Gulf of Mexico is ment of faults distributed widely over parts of western thought to be extensional, and the Yucatan Basin a complex Guatemala, southwestern Mexico and the Gulf of Tehuan- zone of strike-slip faults. The active plate boundary zone 66 tepec. Molnar and Sykes studied earthquake focal mecha- across Guatemala is dominated by the Polochic and Mo-

265 BURKE BURKART

Figure 15.1. Index map of Central America and southern Mexico showing political boundaries of Mexico, Belize, Guatemala, El Salvador (E.S.), Honduras and Nicaragua, the states of southern Mexico and El Peten of Guatemala.

tagua faults, which are both earthquake faults with sinistral tagua River Valley of Guatemala as well as the ophiolite displacement The tectonic significance of the Jocotan- sequences of that region will also be discussed separately. Chamelecon fault, an apparently older fault in this system, Rocks of the Motagua suture zone are discussed in some is controversial. In Guatemala it is the northern boundary of detail by Donnelly et al.34 extensional terrane represented by grabens whose orienta- The composite Mesozoic and Cenozoic sedimentary tions are approximately normal to the fault. section on the northern margin of the Maya block is over The Middle America trench runs parallel to the south- 18,000 m thick. Although this aggregate thickness will ern coast of Mexico and the entire coast of Central America, probably not exist at any one place, it contrasts sharply with marking the zone of northeast-subduction of the Cocos Plate similar sections on the adjacent Chortis and Guerrero beneath the North American and Caribbean plates (Fig. blocks. The thickest Mesozoic -Cenozoic section on the 15.2). Shallow and intermediate earthquakes define the Be- Guerrero block west of the Orizaba fault zone (Fig. 15.2) is nioff zone, revealing that Cocos Plate subduction is at a about 4,000 m, whereas the aggregate thickness of the shallower angle (15°) beneath the southern coast of Mexico Mesozoic -Cenozoic section is about 8,000 m on the Chortis than beneath northern Central America (21°)24. This transi- block in Honduras34,41,57. Aii additional contrast is a section of tion occurs near the central Gulf of Tehuantepec where a more than 3,000 m of Carboniferous -Permian sedimentary northeast-trending, transverse break in the descending plate strata on the Maya block and the apparent absence of is believed to be located. unmetamorphosed Palaeozoic strata on the Chortis block. A probable Palaeozoic age for a thick section of pre-Creta- ceous phyllites, minor marbles and quartzites, exposed on STRATIGRAPHlC SEQUENCE the Chortis block is, as yet, unproven. Palaeozoic sedimentary rocks on the Guerrero block are limited to a few small Stratigraphy of the Maya and Chortis blocks is quite differ - graben basins where thin sequences of Cambrian, Ordovi- ent for rocks older than Miocene. Each block will therefore cian, and Carboniferous strata occur, and to larger basins be discussed separately. Stratigraphic sequences of the Mo- where the maximum thickness of Carboniferous -Permian

266

Northern Central America

Figure 15.2. Major plate tectonic elements of Central America. The Chortis block is the western margin of the Caribbean Plate. Cocos Plate subduction along the Middle America Trench occurs along an east-dipping megathrust The Chortis, Maya (Yucatan) and Guerrero blocks are separated by generalized strike-slip faults indicated by heavy lines. The Maya block has independent motion and is not part of the North American plate as often assumed. Filled circles represent Quaternary volcanoes (not all are represented along Mexican Volcanic belt and Central American Volcanic front). clastics reaches 600 m57. On the basis of its great thickness K-Ar, and Rb-Sr ages ranging from 850 to 1,110 Ma, have of sedimentary section the northern margin of the Maya been considered time-equivalent to Grenville-age metamor- block is exceptional. phic rocks of the Appalachians and Canada39,40. There are no verified Precambrian rocks in the basement complex of Guatemala MAYA BLOCK Pb-U ages of 1,075 Ma and 345 Ma (Proterozoic and Lower Carboniferous) for the Rabinal granite in northern Basement Rocks Guatemala (15°N 90°30'W) were obtained from intersec - Uplifted basement rocks of the Chiapas massif and tions of isotopic ratios and the concordia line. The younger nuclear Central America form an arcuate southern border to age appears to be that of pluton emplacement, whereas the the Maya block, extending from the western Gulf of Tehuan- older may be the age of inherited zircons from the older tepec across northern Guatemala to the Caribbean Sea The metamorphic sequence34,42,60. The Rabinal granite may oldest known rocks of the Maya block are Precambrian correlate with the 'old granite' 59 in the northwest Cuchu- schists and gneisses along the northwestern margin of the matanes Mountains and with the Mountain Pine Ridge Chiapas massif near the town of Arriaga (Fig. 15.4; at 16.5° pluton of Belize, with an age of 336 Ma3. N, 94° W) in the state of Chiapas, Mexico. The Arriaga Damon et al.23 analyzed 10 biotite granite samples with sequence, which is intruded by granites with ages of 702 Ma 21 an apparent Rb-Sr age of 256 ± 10 Ma (Upper Permian), and 780 Ma, was regarded by de Cserna as an extension representing five localities along the northwest part of the of the Precambrian terrane composed of graphitic schists Chiapas massif. They believe this to be a single batholith and gneisses with minor marbles found in the state of and cite evidence that it extends to the western side of the Oaxaca, Mexico. The Oaxaca metamorphics, with Pb-alpha, Gulf of Tehuantepec in the State of Oaxaca. Four Triassic

267 BURKE BURKART

Figure 15.3. (opposite, top) Major structures of northern Central America (from Case and Holcombe16, Burkart and Self13, Peterson 70, Donnelly et al.34 and Home et al.49; unpublished field data for the Orizaba fault zone by B. Burkart and G. Moreno). Filled circles are Quaternary volcanoes. F.Z. = fault zone. Dot pattern outlines margins of deep Tertiary basins of Veracruz, Tabasco, Chiapas and Campeche believed to represent a diffuse exten-sional boundary at the trailing edge of the Maya block 12. The Orizaba fault zone delineates the west margin of the Maya block; it merges with left-lateral strike-slip faults crossing Guatemala and possibly with mapped right-lateral faults sub-parallel to the Pacific coast of Guatemala13. Motagua and Polochic faults tie into the Swan fracture zone in Gulf of Honduras. Extensional tectonics related to strike-slip faulting is prevalent south and east of Mo- tagua and Jocotan fault zones. Right-lateral offset in the Neogene across the Guayape F.Z. of Honduras; earlier displacement believed to be left-lateral38,43.

Figure 15.4. (opposite, bottom) Geologic map of northern Central America from the Isthmus of Tehuantepec to northwestern Nicaragua. Heavy lines indicate mapped faults, with reverse faults indicated by barbs and normal faults by lines pointing to downthrown block. Strike-slip fault offset direction indicated by single arrow. Cabanas fault is just to the south of the Motagua fault in the Motagua River Valley. Light lineations are from Landsat imagery or aerial photographs. Palaeozoic rocks are exposed in the core of the Comalapa anticlinorium (see text), which crosses the frontier between Guatemala and Mexico just south of Angostura Reservoir. Filled circles are Quaternary volcanoes. Map symbols for lithologies are as follows:

Igneous rocks Ti Tertiary intrusions. U Upper Cretaceous (Campanian) ultramafic rocks, primarily serpentinites. UKg Upper Cretaceous granitoids of the Soconuzco and Santa Maria Intrusions of west Guatemala and south Chiapas. Ki Cretaceous intrusions. Pzi Palaeozoic intrusive rocks. I Intrusive rocks, age unspecified.

Metamorphic Rocks Pzm Palaeozoic metamorphic rocks. Includes Chuacus Group metamorphics in Guatemala and Chiapas, San Diego Phyllite in east Guatemala and west Honduras, and Peten Formation in Honduras. Map pattern indicates general direction of metamorphic foliation. Pern Precambrian metamorphic rocks. Grenvillian gneisses and schists.

Sedimentary Rocks TK Cretaceous and Tertiary clastics. In Guatemala and adjacent Chiapas, the Ocozocuautla and Sepur Formations. In Honduras, the Cenomanian to Oligocene Valle de Angeles Group. K Cretaceous marine sedimentary rocks. In northwest Guatemala, the Ixcoy Formation; in central and east Guatemala the Coban-Campur Formation; in Chiapas, the Sierra Madre Limestone and various Upper Cretaceous limestone and clastic units. In east Guatemala and Honduras, the Atima Limestone of the Yojoa Group.

JK Jurassic to Lower Cretaceous Todos Santos terrigenous clastics of Guatemala and southern Mexico. J El Plan Formation and possibly the Honduras Group of Honduras and east Guatemala. Mz Undifferentiated Mesozoic marine and continental sedimentary rocks. Pz Palaeozoic rocks of the Santa Rosa Group of the Maya block.

268 Northern Central America

269 BURKE BURKART

ages between 226 Ma and 237 Ma were calculated for the the Polochic fault, in the core of the Cuchumatanes dome, Hummingbird and Sapote granites of Belize by Donnelly et semi-pelitic and quartzo-feldspathic gneisses, schists, and al.34, using original data of Bateson and Hall3 and recent minor amphibolites have been correlated with the Chuacus decay constants. Donnelly et al.34 reported a Rb-Sr age of Group59. Palinspastic reconstruction (Fig. 15.5) which re- 227 Ma and younger ages of 212 Ma and 213 Ma by Ar39/40 verses 130 km of left-lateral slip across the Polochic fault for the Matanzas granite of central Guatemala (15° 7'N, 90° aligns Palaeozoic metamorphic rocks of the Chiapas massif 12'W). These correlate well with a 238 Ma thermal event with the western Chuacus Group south of the fault to pro- determined by Ar39/40 analysis of amphibolite from the duce a continuous N 60° W orientation of foliation direc - Chuacus Group34. tions8,9. Metamorphic basement rocks of the Maya Marcus59 considered the synorogenic Tenam-Poxlac Mountains of Belize, known as the Maya Series 31, correlate granites of the northwest Cuchumatanes Mountains to be in age and lithology with the Chuacus Group, de Cserna21 post-Palaeozoic and pre- Jurassic in age. A K-Ar age of 196 noted that in Oaxaca, Mexico, Grenvillian (Precambrian) Ma (Lower Jurassic) was reported55 for this granite. Identi- basement rocks are overlain by greenschist and lower am- cal K-Ar ages of 191 ± 4 Ma were reported23 for two phibolite facies rocks whose protoliths were igneous and intrusions from the state of Oaxaca on the Guerrero block: sedimentary rocks of probable Ordovician origin. He attrib- a biotite-quartz-monzanite porphyry (17° 15.8'N, 95° uted metamorphism of the younger sequence to post-Silu- 15.4’W) and a biotote diorite (17° 13'N, 95° 14.8'W). rian orogeny, citing evidence for westward thrusting of an Damon et al.23 reported K-Ar ages from three granite allochthon of these younger schists over the Precambrian and granodiorite samples from near the village of Albino basement sequence. Even though the term Chuacus Group Com) (15° 45'N, 92° 43'W) to be in the range of 170-174 has not been employed outside of Guatemala for the younger ± 3-4 Ma (middle Jurassic). These rocks are cut by andesitic of the two metamorphic sequences of the Maya block, dykes with ages of 141 ± 3 Ma (Neocomian or early Creta- correlative metamorphic rocks are found in Chiapas, and ceous). perhaps on the Guerrero block in Oaxaca. Palaeozoic metamorphic rocks of the Chuacus Group Metamorphic rocks form at least part of the subsurface are exposed in the Sierra de Chuacus between the Polochic basement of the Yucatan Peninsula, as schists and gneisses and Motagua fault zones of central and eastern Guatemala, have been penetrated in petroleum exploration wells in and along the margins of the Chiapas massif (Fig. 15.4). Yucatan and Belize. No correlations have been established 34 Protoliths of this sequence were mainly sedimentary rocks with known sequences . whose metamorphism, according to McBirney60 and McBirney and Bass61, was in the late Devonian. Most of the Soconuzco-Santa Maria Intrusive Rocks rocks of the Chuacus Group are of greenschist facies, con- The Soconuzco Intrusions consist of six or more calc- taining white mica, garnet, epidote, chloritized biotite, al- alkaline granite, granodiorite, monzonite and tonalite plu- bite, sphene and oxides. However, McBirney60 mapped tons that are northwest-trending in the Sierra de Soconuzco almandine amphibolite, and a retrograde metamorphic fa- of the Chiapas massif (Figs 15.4, 15.5). Some of these cies near zones of shearing that lacked garnet and frequently plutons have intruded the Todos Santos Formation and are biotite. Kesler et al.52 and Kesler51 employed the term unconformably overlain by Miocene ignimbrites, making ‘western Chuacus Group’ for a metasedimentary sequence them post-earliest Cretaceous and pre-Miocene. Burkart et 11 of muscovite schists, banded gneisses, and minor marbles al. reported a K-Ar age of 68.4 Ma (Maastrichtian) for the and amphibolites south of the Polochic fault in western southernmost of these plutons in Chiapas, the Motozintla Guatemala It is divided into a muscovite-rich unit, a banded Granite. Across the Polochic fault, 130 km to the east in gneiss and an undivided unit (partly of volcanic protolith), northwestern Guatemala, calc-alkaline granodiorites and each of which represents an original sedimentary lithology granites of the Santa Maria intrusive complex intruded from a granitic source. This sequence forms a southeast- Upper Palaeozoic sedimentary rocks. The Metal Mining 63 plunging anticlinorium cored with metaigneous rocks, gra- Agency of Japan reported K-Ar ages for six granitoid nodiorite and monzonite. The westernmost exposures of the rocks in the Santa Maria complex. Two quartz monzonites Chuacus Group south of the Polochic fault in Chiapas, provide an Aptian and Neocomian, or early Cretaceous, age Mexico are quartzo-feldspathic chlorite augengneiss, (135 Ma, 117 Ma), while three granodiorites indicate an metaquartzite, black phyllite, mica schist and metaigneous Upper Cretaceous age interval from Santonian to Maas- rocks 11. Quartz-feldspar -biotite gneisses and augengneisses trichtian (85, 83, 74 Ma). K-Ar ages of Paleocene-Eocene with igneous protoliths in the composition range of grano- (62 and 58 Ma) are reported for two rhyolite sill and dyke diorite to diorite are found north of the Polochic fault along samples. the Pacific margin of the Chiapas massif51. Also north of On the 130 km sinistral palinspastic reconstruction

270 Northern Central America

Figure 15.5. Palinspastic reconstruction reversing 130 km of left-lateral slip across the Polochic fault Mid- Miocene restoration showing the proto-Polochic fault. Motagua and Jocotan faults traces are as they exist today. The Tonala fault83 is a part of the Orizaba fault zone. The Arista fault83 is an extension of the Polochic fault zone in the Gulf of Tehuantepec. All fault traces are dashed. Reconstruction juxtaposes Chiapas massif to crystalline rocks of western Guatemala (Nuclear Central America). PC = Precambrian metamorphic rocks; Pzi = Palaeozoic intrusive rocks; Pzm = Palaeozoic metamorphic rocks (wavy pattern); dot pattern = late Palaeozoic Santa Rosa Group; UKg = Upper Cretaceous granitoid rocks of Soconuzco intrusions and Santa Maria intrusions (plus pattern); question mark suggests the possibility that Jocotan-Chamelecon or Motagua fault zones may have connected with the Orizaba fault zone at this time.

across the Polochic fault the Soconuzco intrusions form a southern limb and nose of the Comalapa anticlinorium (Fig. continuous northwest-trending belt with the Santa Maria 15.4), where 6,300 m of section (some very likely repeated) Granites (Fig. 15.5). There are other Upper Cretaceous -Ter- is increasingly metamorphosed to garnet schist towards the tiary intrusions of similar composition range along the plate base. The Lower Santa Rosa Formation consists of phyllitic boundary zone in northern Central America (see below). shales and anhydrite whose identifiable fauna has a Lower Carboniferous to Upper Permian range. This sequence is Sedimentary Rocks discordantly overlain by the unmetamorphosed Santa Rosa Lower Carboniferous rocks, the oldest known sedimen- Formation, composed of Desmoinsian-Upper Virgilian tary rocks of the Maya block, are reported from two regions. (Upper Carboniferous) sandstones and shales, interbedded The first is the northern Cuchumatanes Mountains, where with fossiliferous limestones 46,57. the Cantel Sequence (Fig. 15.6), is believed to be upper The Santa Rosa Group of northwestern Guatemala Lower Carboniferous to lower Upper Carboniferous 55. The overlies the Lower Carboniferous Cantel strata in the Cantel Sequence, which unconformably overlies Chuacus Cuchumatanes Mountains and has been divided into four Group schists, is separated by unconformity from the over- units: the Sacapulas, Tactic, and Esperanza Formations, and lying Santa Rosa Group. It consists of micaceous sandstones Chochal Limestone (Fig. 15.6). The Sacapulas and Tactic and black shales interbedded with volcaniclastic layers with Formations may be in part Upper Carboniferous, but the volcanic glass shards and radiolarian cherts intruded by Esperanza Formation, which consists of interbedded lime- dolerite sills55. The second reported occurrence of Lower stone and shales, and the Chochal Limestone, are Wolfcam- Carboniferous sedimentary rocks, referred to as the Lower pian and Leonardian, respectively (Lower and Middle Santa Rosa Formation, is found in Chiapas, Mexico on the Permian). Along the southern front of the Cuchumatanes

271 BURKE BURKART

Figure 15.6. Correlation chart for late Palaeozoic stratigraphic units: (1) southern Chiapas, including southern part of Chiapas fold belt 46,57,88; (2) southern Cuchumatanes Mountains of northwest Guatemala18; (3) Northern Cuchumatanes Mountains of northwest Guatemala55; and (4) Maya Mountains of Belize45. The Lower Santa Rosa Formation, Cantel Sequence and Maya Mountains sequence containing the Bladen Volcanic Member may be stratigraphically equivalent. If so, the Bladen Volcanic Member is probably not a part of the Santa Rosa Group.

Mountains in northwestern Guatemala the Sacapulas For- found them to be absent in northern Yucatan. The entire mation is overlain by shales of the Tactic Formation, above Palaeozoic section is missing in some exploration wells in which interbedded limestone and shale of the Esperanza the central part of the Yucatan Peninsula. Formation is capped by a massive dolomitic limestone with In the Maya Mountains of Belize, Dixon31 defined two shale, the Chochal Limestone. Northward across the Cuchu- Upper Palaeozoic units, the Maya Formation and the discor- matanes Mountains in northwestern Guatemala, interbed- dant, overlying Macal Sandstone. Upper Carboniferous to ded carbonates of the Esperanza Formation are replaced by Permian fossils have been identified in the Macal Sand- shales, and the distinction between Tactic and Esperanza stone, which was later correlated with and placed in the Formations cannot be made. This, along with the existence Santa Rosa Group by Bateson2. Hall and Bateson45 defined of an oolitic facies in the Chochal Limestone in the northern the Bladen Volcanic Member within the clastic sequence. 55 Cuchumatanes , is suggestive of a northward shoaling, and Anderson et al.1 correlated the Chicol and Sacapulas For- is consistent with the argument of Hall and Bateson that mations of northern Guatemala with the Bladen Volcanic the Permian shoreline was not far north of Maya Mountains Member. Carbonates of the Maya Mountains are thinner 27 exposures in Belize. Dengo reported Upper Carbonifer- than the Chochal Formation of Guatemala, but are of equiva- ous-Permian age strata in wells in southern Yucatan, yet lent Permian age.

272

Northern Central America

Figure 15.7. Correlation chart for Mesozoic and Cenozoic stratigraphic units of the Maya block and the Motagua Valley (after Donnelly et al.34). Additional information includes radiometric ages of Santa Maria and Soconuzco granodiorites in west Guatemala and Chiapas, respectively (filled circles). Time correlation of Ocozocuautla and Sepur Formation clastics containing probable volcanic equivalents of granodiorites is noted, as is time of ophiolite emplacement in eastern Guatemala.

In southern Chiapas the name Upper Santa Rosa For- 1,200 m in thickness, thinning to a few 10s of m across the mation is equivalent to Sacapulas and Tactic Formations of basement horst of the Tenam-Poxlac uplift. adjacent Guatemala. The Upper Santa Rosa Formation was Shales and carbonates of the San Ricardo Formation of 58 found by Malpica-Cruz to be Upper Carboniferous on the Chiapas, Mexico were deposited during marine transgres- basis of algae of the genus Komia. Upper Permian sedimen- sion across active extensional basins that may have been tary rocks are not known on the Maya block, evidently related to a continuation of the opening of the Gulf of because late Palaeozoic orogenesis (see below) had begun Mexico initiated in the late Jurassic5,6. by that time. The Great Cretaceous Carbonate Event Jurassic-Cretaceous Red Bed Basins and Evaporites Steele87 and Waite92 measured a 5,000 m sequence of Mesozoic salt deposits are distributed north of a region platform limestones and dolomites deposited from Middle bounded by the western margin of the Isthmus of Tehuan- Aptian to Middle Santonian in Chiapas, Mexico. This se- tepec and Chiapas-Guatemala massif as far west as central quence is referred to as the Sierra Madre Limestone in Guatemala, but are absent in eastern Guatemala, the Peten, Chiapas, Mexico, the Orizaba Limestone in Pueblaand Vera Belize, and approximately the eastern third of the Yucatan Cruz states of Mexico, the Ixcoy Limestone in northwestern 4,70,89 Peninsula . These evaporites have given rise to salt Guatemala, and the Coban and Campur Limestones in cen- domes and salt massifs in the Isthmian Basin of Veracruz tral and eastern Guatemala and the subsurface of the Peten. and offshore Gulf of Mexico. Whereas earliest salt was of Although some have been explained as solution breccias, 4 Jurassic (pre-Oxfordian) age, Bishop suggested salt was other notably thick, massive informational breccias and deposited throughout the late Jurassic and most of the Cre- conglomerates in the Ixcoy and Coban Limestones of Gua- taceous somewhere within the region described above. temala are of possible tectonic origin1,7. In Chiapas this Jurassic evaporites in Chiapas, Mexico and northern great carbonate section overlies the San Ricardo Formation Guatemala were followed by the Todos Santos and San or is unconformable on basement rock of the Chiapas mas- Ricardo Formations, which comprise an Upper Jurassic - sif. It extends across Guatemala into the Caribbean and into lowermost Cretaceous siliciclastic -carbonate sequence de- the subsurface of Tabasco and Chiapas states of Mexico 5,6 posited during rift basin development . The Todos Santos where it is a major petroleum reservoir rock. In the Reforma Formation consists mainly of fluvial deposits in northwest- and offshore Campeche districts the Sierra Madre Lime- trending graben basins (Fig. 15.7). In the Cuchumatanes stone is productive from carbonate reef, reef talus, karstic Mountains of northwestern Guatemala the sequence reaches limestone, and dolomitic limestone, with enhanced reser-

273 BURKE BURKART

voir porosity from fracturing 70. sedimentary basins are complex, having formed in associa- 68 Arc-related flysch of the Ocozocuautla-Sepur Formations tion with strike-slip and local reverse faulting . Miocene The Cretaceous carbonate sequence was succeeded by sedimentary rocks are mainly marine clastics in the basins deposition of siliciclastics of the Ocozocuautla Formation of Veracruz, Tabasco and northern Chiapas, but grow more in Chiapas17, and the Sepur Group and Toledo Formation in calcareous eastward onto the Yucatan Peninsula. Tertiary 70 Guatemala and Belize, respectively30,34,80,90. The Cam- carbonates extend across the entire Yucatan platform . panian-Maastrichtian Sepur Formation of southeastern Several Miocene conglomerates of local extent occur in Guatemala was interpreted by Rosenfeld80 as a submarine the Peten and along the Caribbean coast. The Rio Dulce fan deposit containing ophiolitic debris with up to 1 km or Formation, a Lower to Middle Miocene fossiliferous lime- larger serpentinite olistoliths, and bearing a significant com- stone of up to 1,000 m thickness, is found along the coast 34 ponent of plutonic and volcanic detritus with calc -alkaline and near Lago de Izabal . The Pliocene Herreria Formation affinities. Rosenfeld80 commented as follows on the prove- is composed of claystones, marls and limestones with occa- nance of these clasts: "The igneous suite in the conglomer- sional lignites, cropping out around the eastern part of Lago ate therefore consists of subaerially erupted lavas and de Izabal and along the coast of Guatemala and southern pyroclastics, and hypabyssal acid plutonic rocks typically Belize. associated with a partially denuded volcanic arc." Wilson98 Little has been published on Tertiary continental and and Rosenfeld80 concurred that the sediment source for the marine basins along the Caribbean margin of eastern Gua- Sepur flysch deposits was from the south, but Rosenfeld80 temala and Belize, but oil company exploration reports could find no suitable source for the igneous components. suggest many extensional basins are developed adjacent to The Oxec ophiolite, exposed just north of Lago de Izabal in a widespread set of northnortheast-trending left-lateral strike-slip faults, some with significant normal separation. the Sierra de Santa Cruz (Fig. 15.4), is an Upper Campanian 4 slide-mass that separates upper from lower Sepur Group Bishop described block-faulted topography along the east- clastics34,80. In southern Belize flysch deposits of the ern coast of the Yucatan Peninsula extending from Quintana equivalent Toledo Formation contain fossils as young as Roo into Belize. Downfaulting to the east along northnorth- early Eocene. The correlative Ocozocuautla Formation in east-trending normal faults is believed to have begun in the Chiapas is a sequence of shallow shelf deposits of conglom- early Tertiary and to be still active. Sediment thicknesses of erate, sandstone, shales, and limestone, with abundant inter- at least 5,000 m are present in the Lake Izabal half graben, bedded volcanic ash derived from a late Cretaceous formed by left-lateral offset across the Polochic fault. The magmatic arc to the south. Armas Formation, consisting of about 1,400 m of conglomerates, sands, red beds and deltaic sedimentary rocks, is Tertiary sedimentary units located within a fault-bounded depression in the Motagua Tertiary sedimentary rocks reach an estimated thick- River Valley near the Caribbean coast. ness of between 8,000 and 10,000 m along coastal regions of Veracruz and Tabasco, and offshore in the Gulf of Mex- ico70. These mostly fine-grained clastic marine sequences MOTAGUA FAULT ZONE are essentially absent west of the Isthmus of Tehuantepec, and in the Chiapas fold and fault belt are preserved only The El Tambor Group, an ophiolite assemblage in the Mo- where downfolded or faulted against Sierra Madre Lime- tagua River Valley and adjacent ranges of central and east- stone. Three major Tertiary block-faulted basins are found ern Guatemala, is compelling evidence for late Cretaceous collision between the Maya and Chortis blocks. Donnelly et in the Gulf of Mexico coastal region, the Veracruz, Comal- 34 calco, and Macuspana basins, the latter two separated by a al. summarized the information on ophiolites of the Mo- horst upon which a major petroleum region, the Reforma tagua suture zone, which consist of serpentinite, pillow trend, is situated (Fig. 15.3). Over 4,000 m of Paleogene basalts, interbedded cherts and basalts, wackes, phyllites marine clastics are associated with turbidites in the centres and schists, with local exposures of gabbros, plagiogranites, of these coastal basins. Original thicknesses of up to 5,000 slightly serpentinized peridotites and pelagic limestones. m of continental and marine sediments were deposited in the Structural evidence supporting collision includes north-ver - Paleogene in a narrow (circa 70 km wide), elongate basin gent thrust sheets to the north of the Motagua Valley and along the Chiapas fold belt and across northern Guatemala south-vergent thrusts to the south. Abundant amphibolites to the Caribbean70. Over much of the Chiapas-Guatemala and rare eclogites are metamorphic equivalents that occur in fold belt the Tertiary sedimentary section is absent because regions adjacent to the suture zone. An early Cretaceous to of early to middle Cenozoic uplift and erosion. Tertiary Cenomanian age has been assigned to basalts of the El Tambor Group, while a late Campanian age has been given

274 Northern Central America

to the time of collision along the Motagua suture zone. Precambrian or Cambrian48-50. The correlative Las Ovejas In northeastern Guatemala an allochthonous volcani- Complex, located along the southern border of the Motagua clastic wacke sequence of Aptian-Albian age, called the Valley of Guatemala, is composed of quartzofeldspathic Tzumuy Formation, is found north of Lake Izabal80. This gneisses, two-mica schists, amphibolites and marbles. Silli- sequence is comprised of coarse-grained beds containing manite is common, while andalusite and staurolite are found andesite detritus interbedded with fine-grained sedimentary locally. rocks deposited on the upper pillow basalts of the Sierra de A younger metamorphic sequence of phyllites, gra- Santa Cruz ophiolite, and is believed to represent the initial phitic schists, and minor marbles and quartzites is called the age of obducted ocean crust. Rosenfeld80 suggested possible San Diego Phyllite in eastern Guatemala, and the Peten and correlation with the Devil's Racecourse Formation (pre-late Cacaguapa Formations in central Honduras. The sequence Albian) of Jamaica, and with similar units of Hispaniola and is unconformable with the overlying Mesozoic sedimentary Puerto Rico. rocks of southeastern Guatemala and central Honduras, and Continental red beds of the Eocene Subinal Formation in the Sierra de Omoa it is unconformable above the older occur in two elongate structural basins in the Motagua River metamorphic sequence10,14,36,48,50. These rocks are bracketed Valley. There the Subinal Formation has been divided into within a range from 140 to 305 Ma. two units, a lower quartz pebble conglomerate and an upper A series of granitoid plutons occurs in the northern conglomeratic red sandstone displaying typical fluvial marginal complex intruding the Las Ovejas Complex and 34 structures . Rare occurrences of the brackish-water gastro- San Diego Phyllite. Home et al.48 described Laramide plu- pod Lagunitis, of Eocene age, have been found in both upper tons of tonalite and granodiorite in the Sierra de Omoa of and lower units. The Subinal Formation has been extended northwestern Honduras, Clemons and Long19 described the in name to Eocene red beds in southeastern Guatemala, Chiquimula Pluton of late Cretaceous age (84, 95 Ma), where up to 3000 m of andesitic conglomerates, fluvial composed of granite, granodiorite, diorite and gabbro, found sandstones and siltstones, and minor andesite flows overlain just south of the Motagua fault zone in eastern Guatemala. 10,33,44 by volcanic arenites fill a northeast-trending graben . Donnelly et al.34 believed that this is possibly a composite A K-Ar age of 42 Ma (late Eocene) has been determined for pluton of later Cretaceous-early Tertiary age. Ritchie and andesite clasts in the lower conglomeratic sequence. McDowell77 described a granitoid pluton north of Guate- Two significant points relate to the structural occur- mala City of late Cretaceous age. rences of the Eocene red beds of Guatemala: they are not Very few field studies have been conducted in the present on the Maya block; they are found in parallel block- remote eastern area of Honduras and northern Nicaragua faulted basins bounded by segments of the Motagua and called Terra Incognita by Home et al.49, where widespread Cabanas faults (Fig. 15.4), and near Quezaltepeque in south- exposures of metamorphic rocks of two sequences have eastern Guatemala, by an unnamed fault parallel to the been reported. Zoppis Bracci99 defined the Palacaguina Motagua fault trend. The depositional axis of this southern Formation, a greenschist facies association of phyllites, basin is also parallel to the Motagua trend. Clearly these schists, and minor quartzite and marble beds, apparently were intermontane basins bounded by active faults with 33 correlative with the San Diego Phyllites and Cacaguapa considerable dip-slip components . It remains to be proven Formation. Engels35 described a greenschist facies assem- that there was a significant strike-slip component to the blage of metamorphosed pillow basalt, volcanic ash and faulting that formed these basins. On the basis of ages, dolerite exposed in northern Nicaragua, which he consid- geometries and positions in the plate boundary zone, it is ered to be younger than the Palacaguina Formation. Mills this author's opinion that they are strike-slip basins and that and Hugh64 regarded the volcanic sequence to be Creta- the Maya-Chortis transform boundary was active in the ceous and lower Tertiary. Eocene. A Rb-Sr age of l40±15 Ma(early Cretaceous) has been reported49 for the Dipilto batholith, a calc-alkaline CHORTIS BLOCK adamel-lite to tonalite intrusion into the Palacaguina Formation. This is one of many similar plutons distributed in an east-northeast trend in northwestern Nicaragua. Basement Rocks The oldest known rock of the Chortis block is in a Sedimentary Rocks basement ridge complex cropping out along the northern Palaeozoic sedimentary rocks have not been reported margin. A metaigneous sequence of granodiorite, gneiss and south of the Motagua fault zone, and Mesozoic sedimentary tonalite hornfels in the Sierra de Omoa of western Honduras sections are quite different north and south of the fault. The is pre-Upper Carboniferous and possibly as old as upper El Plan Formation of central Honduras, which is probably

275 BURKE BURKART

Figure 15.8. Late Cretaceous island arc terrane in Chiapas, Mexico. Section A-A' begins in Gulf of Tehuantepec Palaeozoic metamorphics (wavy pattern) overlain by Carboniferous though Upper Cretaceous sedimentary rocks (dot pattern-, well data reported by Sanchez-Barreda83). Upper Cretaceous (UK) volcanics are hypothetical as volcanic equivalents of UK granodiorites are known only in UK sedimentary rocks (Gulf of Tehuantepec forearc basin; fine- to coarse-gained clastic sequences with volcanic ash and conglomerates of tonalite, dacite and porphyritic basalt pebbles *)• UK continental back arc basin; Ocozocuautia Fonnation shallow-water clastics with serpentinite grains, containing ash beds and limestone. Collision of Chortis and Maya blocks32 required subduction and, probably, arc magmatism as suggested here. Concordia fault zone with high angle reverse faults67 is margin of uplifted massif A similar cross section (south to north) for eastern Guatemala would terminate on the north in the deeper Sepur clastic basin. There the arc intrusions are missing today, possibly by displacement across postulated transform faults. the oldest sedimentary unit on the Chortis block, consists of nated horizons and skeletal sand shoals. Deep-water sandstone, siltstone and shale ranging from a few to several 10 49 facies are reported from southeastern Guatemala and hundred metres in thickness. Home et al. considered that central Honduras 98. the El Plan Formation may have been deposited during the Tertiary block faulting has, no doubt, complicated the Jurassic in a terrigenous and shallow shelf environment Part study of Mesozoic stratigraphy of the Chortis block, but the of the El Plan Formation may correlate with the Honduras stratigraphy itself is inherently complex, with large thick- Group, a Jurassic to possibly lowest Cretaceous sequence of ness and facies changes occurring in relatively short dis- siliciclastics found in southeastern Guatemala and Hondu- 49,76 tances. Depositional basins appear to have been technically ras . The Honduras Group is unconformable on base- active during the time of deposition of the Atima Formation ment rocks and conformable with overlying limestones of which has thicknesses ranging from 1,700 m to as thin as the Atima Formation. 100 m in areas adjacent to old basement highs10,49. In Limestones of the Aptian-Albian (Lower Cretaceous) southeastern Guatemala just north of the Jocotan fault, only Atima Formation are found from southeastern Guatemala to 300 m of Atima Limestone is found unconformably overly- central Honduras. Recent revision of stratigraphic nomen- ing basement rock of the northern marginal complex. Adja- clature49 redefined the Atima Formation as the only unit of 65 cent and to the south of the Jocotan fault the section the Yojoa Group . The Atima Formation is dominated by increases to 1,290 m of thin-bedded Albian limestone sepa- massive-bedded, dark gray biomicrite, of a shallow car- rated by 200 m of fanglomerate from 430 m of massive-bed- bonate shelf environment broken by small, rudist-domi- ded limestone. Donnelly et al.33 suggested significant

276 Northern Central America

dip-slip movement up-to-the-north across the Jocotan fault adjacent and to the south of the Jocotan fault, and pull-apart in the Albian. This relationship has been noted by Home et basins along the Motagua fault zone. Silicic volcanism al.50 across the Chamelecon fault, a continuation of the accompanied extension, with ignimbrites and volcaniclastic Jocotan fault in Honduras. Just 30 km south, near the Gua- infilling of the resulting basins13,49,96. Gordon43 reported a temala-El Salvador border, the combined Atima Limestone number of extensional basins located in central and eastern section thins to 950 m, with 400 m of thin-bedded limestone Honduras, mostly associated with right-lateral faults. overlain by 550 m of massive-bedded limestone. Burkart and Self13 relate these basins to extension south Conformably overlying the Atima Limestone is the of the Jocotan fault. Valle de Angeles Group, which consists of two unnamed, red siliciclastic formations and an intervening marine lime- Central American volcanic arc stone, the Jaitiqui Limestone of western Honduras or Es- The Quaternary volcanic arc of Central America ex- quias Formation of central Honduras36,37,50,65,97. The lower tends 1,100 km northwestward along the Pacific coast from conglomeratic sandstone ranges from about 1,000 m to Panama to the Mexico-Guatemala border. Volcanic geology 2,000 m in thickness and appears to have been deposited in of Central America has been summarized by Weyl95 and an intermontane graben basin. The clastics grade into and Carr and Stoiber15, who reported more than 580 mainly are conformable with the Cenomanian Jaitiqui/Esquias andesitic volcanoes along this essentially continuous belt, limestones, whose thicknesses range from 200 to 400 m. The about 40 of which are significant, independent edifices. thickness of upper red bed sequence of fine-grained clastic Central American volcanoes 15 have produced 16 cubic km rocks ranges from 0 to 1,000 m. Volcanic detritus is found of volcanic products since the year 1680, in contrast to the throughout the unit with nodular gypsum present in the Lesser Antilles arc, which has produced 1 cubic km along lower part. A transitional to near-shore environment has its 750 km length. Lavas are primarily basalts and andesites, been interpreted for this unit, including deltaic, sabhka, but silicic ash-flow tuffs are particularly abundant in the floodplain and shallow marine facies. The Jaitiqui Lime- volcanic highlands of Guatemala. The Central American stone lies within the Cenomanian-Oligocene time inter- volcanic arc terminates with Volcan Tacana at the Guate- val34. mala-Mexico border, where the Polochic and Motagua fault Although the Valle de Angeles Formation of El Salva- zones merge. Quaternary volcanic centres north of these dor94, and the Totogalpa Formation of northern Nicara- faults are not only farther inland than the Central American gua62,100, have been correlated with the Eocene Subinal arc, but are small and widely scattered22. Displacement Formation of Guatemala, the age of these units is unknown inland of Maya block volcanism from the Central American and the correlation is uncertain 4. The formations are domi- arc may be the result of the shallow subduction angle that nated by continental volcaniclastic conglomerates, sand- begins in this region, with the occurrence of volcanism in stones, shales and water -depos ited silicic ash interbedded places of crustal extension. with andesite flows.

PRE-MESOZOIC HISTORY OF NORTHERN CENTRAL AMERICA LATE TERTIARY VOLCANIC ROCKS Grenvillian gneisses and overlying Chuacus Group schists, Ignimbrites, flows, and volcaniclastic sedimentary rocks the two basement rock units of the Chiapas massif, are Miocene and younger ignimbrites, lava flows, and vol- probably correlative with two sequences of uplifted meta- caniclastic sediments occur on both the Maya and Chortis morphic basement in Oaxaca, where Cserna20 reported a blocks, spanning a region from the State of Oaxaca, Mexico major late Ordovician orogenic event including overthrust- to Costa Rica. These volcanic deposits attain thicknesses of ing. There have been few studies of basement rocks of up to 2,000 m and form the base of the volcanic province of Chiapas and this event has not been confirmed there. North- the western Isthmus of Central America. In eastern Guate- west-trending directions of metamorphic foliation are com- mala and most of Honduras rhyolitic tuffs, volcaniclastic mon to both the Chuacus Group schists of the massif and sedimentary sequences and capping basalts are known as the the younger sequence of metamorphic rocks of Oaxaca. 10 Padre Miguel Group . Ignimbrites of Miocene age of similar A major unconformity exists on the Maya block be- character are found on the Maya block in western Gua- tween Chuacus Group metamorphics and platform clastics temala, along the Sierra de Soconuzco drainage divide and and carbonates of the Lower Carboniferous to Lower Per- as scattered deposits in the State of Oaxaca. Miocene to mian Santa Rosa Group. In Guatemala Upper Carboniferous Holocene pyroclastic flows and terrigenous volcaniclastic siliciclastics with radiolarian cherts and volcaniclastic de- sedimentary rocks partially fill the north-trending grabens posits constitute the oldest sedimentary sequence above this

277 BURKE BURKART

unconformity. The Cantel Sequence of northern Guatemala mid-late Cretaceous) by the great shelf carbonate sequence is Lower Carboniferous, deposited during the postulated late in Mexico from Chiapas northward to offshore Gulf of Palaeozoic convergence of North and South America when Mexico, and in northern Guatemala and Belize. In this time the Maya block may have been situated in the northern Gulf interval thick evaporites extended into the platform interior of Mexico81. It is possible that this sequence represents the across northern Chiapas and the Peten of Guatemala. Car- accretionary prism of an island arc system. bonate buildup on the Yucatan platform continued until the There has been debate over the existence of a major Holocene70. deforniational event in the post-Leonard (Permian) to pre- Structure of the Maya Mountains of Belize, which were late Jurassic time span in northern Central America1. In the initiated in the Cretaceous, is dominantly that of a horst breached Comalapa anticlinorium in Mexico about 100 km block bounded by northeast-trending faults. Dengo and northwest of the frontier, Lopez Ramos56 reported a major Bohnenberger30 considered that the mountains were topog- angular unconformity between Permian limestones and the raphically high in the Cretaceous. Cretaceous carbonates overlying Todos Santos Formation. Webber and Ojeda93 thin toward the Maya Mountains, but Bishop4, in an inter- reported northeast-trending folds and northwest-trending pretation of exploration seismic and well data, determined faults in Permian beds of southeastern Oaxaca that did not that the uplifted region was submerged in the early to middle involve the overlying Mesozoic strata. In eastern Guatemala Cretaceous. Palaeozoic granites, Maya Series metamor- there is a major angular unconformity between the Lower phics (Chuacus Group equivalents) and Santa Rosa Group Permian Chochal Formation and an underlying sandstone sedimentary rocks are exposed on this structure over a wide (Las Escobas beds) that is clear evidence for an early Per - area of east-central Belize (Figs. 15.3, 15.4). This large 34 mian deformational event . Granitoid plutons of Permian- block tilts westward and is known as La Libertad arch in the Triassic age are found in the Chiapas massif and core of the subsurface of the central Peten. 59 Cuchumatanes anticlinorium , yet no angular unconfor- Northeast subduction in the late Cretaceous beneath the mity has been detected between late Palaeozoic sedimentary Chiapas massif and western Guatemala produced the vol- rocks and overlying Mesozoic sedimentary sequences in the canic arc represented by granodiorite plutons that penetrate Comalapa anticlinorium in the Cuchumatanes Mountains of 1,10,11 60 the basement terranes of the Chiapas-Guatemala massif. northwestern Guatemala . McBirney found no con- Intrusions of that age are not present in eastern Guatemala vincing evidence of a mountain-building episode in the north of the Motagua fault zone, but the existence of rocks post-Permian to pre-Jurassic (Todos Santos Formation) of similar ages and compositions on Jamaica and the Cay- time interval in the central Guatemalan Cordillera. The con- man Ridge raises the possibility (as of now unproven) that cordance between Palaeozoic and Mesozoic strata has been these intrusions are displaced terranes from northern Central documented only in two regions that are exactly juxtaposed 9 America. on the 130 km palinspastic reconstruction , suggesting that The great carbonate event ended in southern Chiapas the observation is valid, but for a relatively small area. The and northern Guatemala with outpourings of volcaniclastic late Palaeozoic was certainly a time of major orogeny on the debris into continental back arc basins during the develop- Maya block, because the unconformity is profound, repre- ment of the Soconuzco volcanic arc in late Cretaceous-Pa- senting a time span from middle Permian to late Jurassic leocene (Fig. 15.8). Volcanic debris is found in the during which uplift, igneous intrusion and folding has been Campanian-Eocene Ocozocuatla and Sepur Formations documented. from central Chiapas to eastern Guatemala. The Chiapas -Guatemala fold belt was initiated in the MESOZOIC TECTONIC HISTORY OF late Cretaceous in response to northeast convergence. Don- NORTHERN CENTRAL AMERICA nelly32,34 noted that convergence culminated with obduction of ophiolites and uplift of basement rock on thrust faults Opening of the Gulf of Mexico is thought to have begun in along the Motagua suture zone, interpreting ophiolite em- the Oxfordian (late Jurassic) as the Maya block moved placement as the result of collision of the Chortis and Maya southeastwardly with respect to the North American plate blocks in the Campanian. An alternate hypothesis has been across a transform boundary in much the same position as advanced71,72,74,81 , wherein suturing of a proto-Greater An- the Orizaba fault zone. Transtension across this fault zone tilles volcanic arc to the Maya block resulted from north- may have caused late Jurassic and early Cretaceous rifling eastern movement of the western part of a proto-Caribbean of the Maya block resulting in the salt basins of the south- plate. Later suturing of the eastern part of that arc to the ernmost Gulf of Mexico and terrigenous clastic basins of the Bahama platform would have taken place with continued Todos Santos Formation. The Todos Santos Formation was plate movement. succeeded from the Neocomian through Santonian (early to The metamorphic fabric of the Chuacus Group on the

278 Northern Central America

Maya block is parallel to late Cretaceous compressional Cayman Ridge", an event they believed to be middle to late structures, forming an arcuate bend from the Chiapas massif Eocene in age, that "transferred the Chortis Block and southward across northern Guatemala to the Caribbean51. associated terranes from the North American plate and Near the Motagua suture zone the original structure of the added them to the Caribbean plate". Pindell and Dewey74 Chuacus Group metamorphic rocks has been obscured by postulated that this motion began at 36 Ma (earliest Oligo- deformation that began in late Cretaceous on a north-trending cene). As surmised above, the late Cretaceous volcanic arc congressional axis34. Overprinting has not occurred in the of Chiapas and western Guatemala would be shut off from Chiapas massif or in the core of the Cuchumatanes subduction as Chortis moved adjacent and, indeed, that arc anticlinorium. There the Chuacus Group metamorphic rocks seems to have become inactive in the Paleocene. are penetrated by undeformed plutonic rocks with radiometric ages ranging from late Permian through late Juras- OFFSETS ACROSS THE MAJOR sic 23,57,59. The coincidence of northwest-trending structural STRIKE-SLIP FAULTS fabrics in Chuacus Group metamorphic rocks and Mesozoic rocks of those regions must represent two separate times of Estimates of total slip across the Polochic, Motagua, Jocotan compression along the same direction. Late Cretaceous- or other faults in Guatemala, Honduras, and Chiapas Mexico early Tertiary (Laramide) compressional orientation was have been the subject of much controversy. As stated above, probably normal to the northwest-trending continental mar- Hess and Maxwell47 implied about 600 km of left-lateral gin in Chiapas, the Chiapas massif. offset across postulated Cayman Trough strike-slip faults which they suggested might extend across the Central American isthmus. About 1,100 km of left-lateral slip has ORIGIN OF THE CHORTIS BLOCK been determined by Pindell and Dewey74 from the length of the Cavman Trough. Wadge and Burke91 and Rosencrantz et The origin of the Chortis block has been the subject of much al.79 agreed with an offset of this magnitude. The three 34 speculation. Donnelly et al. listed the following in their major fault zones crossing Guatemala are discussed below review of published hypotheses of suspected early Meso- from the perspective that one or more may be candidates for zoic positions: in the central Gulf of Mexico; against the this major offset. eastern Maya block in the Gulf of Honduras; adjacent to the The Polochic fault trends westward across northern Guerrero block (the south coast of Mexico); off the north- Guatemala, cutting obliquely across northwest-trending west coast of South America; in the Pacific Ocean; and in outcrop belts and structures, and separating terranes of great its present position. The position adjacent to the Guerrero contrast. It is the only one of the three major faults in 29 block appears to be the most favoured at this time. Dengo Guatemala that does not run parallel to regional contacts and 34 and Donnelly et al. noted that Mesozoic sedimentary structure, the major factor in facilitating a match of geologi- rocks are similar on the Chortis block and in the Morelos- cal features across the fault which leads to a measurement Guerrero basin along the south coast of Mexico. The two of 130 km left-lateral displacement9. The fault, which is successions also have comparable thicknesses (see above) essentially west-trending across the isthmus, bends to the and unconformable relationships to basement metamorphic westnorthwest and appears to merge with the Arista fault83 in sequences. the Gulf of Tehuantepec. Thus, contrary to previous Authors who support a Guerrero block origin for Chortis speculation9, it does not cross the Gulf of Tehuantepec to are in general agreement that it has reached its present the Middle America trench (Fig. 15.3). Along the Polochic position by eastward transform fault displacement, but there fault zone there are at least three major fault wedges, is less agreement on the time of initiation of strike-slip bounded by the active fault and by abandoned segments of motion, and whether a Chortis-Maya block collision took the fault11,26. In eastern Guatemala the Polochic fault splays 32 place. Donnelly considered this movement was initiated into three segments (Fig. 15.4). Schwartz et al.86 mapped before 119 Ma (Aptian), with most of it over after Chortis- topographic offsets on the northern splay that runs north of Maya block collision at 65 Ma (end Maastrichtian). Lake Izabal (North Izabal fault9). The central splay, which A drawback to a pre-Aptian emplacement of Chortis is passes beneath sediments of Lake Izabal just southeast of the difficulty in explaining the late Cretaceous volcanic arc the northwestern shore, is visible on multichannel seismic of Chiapas and western Guatemala, because Chortis would profiles across the lake (data from petroleum exploration be in the way of northeast subduction if it were attached to programs). This splay constitutes the major fault boundary 81 the Maya block at that time. Ross and Scotese described to the Lake Izabal half graben, which has major vertical development of the "E-W trending Polochic -Motagua-Jo- separation down to the northwest. Transtension clearly ac- cotan faults with the southern strike-slip margin of the companied left-lateral slip across this segment of the fault,

279 BURKE BURKART

which is probably the splay with the greatest amount of stones. Muehlberger and Ritchie69 believed the Jocotan- strike-slip offset Chamelecon fault to be an inactive segment of the plate The following geological features have been correlated boundary zone across Guatemala, with the Motagua and across the Polochic fault to establish its 130 km offset (Fig. Polochic fault zones being more recently active. Burkart and 15.5): basement terrane of the Chiapas massif and that of Self 13 stressed the importance of the extended terrane south northern Guatemala (nuclear Central America); faults, folds of the Jocotan-Chamelecon fault zone and the scarcity of and outcrop patterns exposed north of the fault in the north- such terrane to the north. They provided a model whereby west-trending Chiapas-Guatemala fold belt, and matching graben basins grow due to rotation and extension of crustal features south of the fault; the Upper Cretaceous Soconuzco blocks around the fault, with left-lateral slip accumulating intrusions north of the fault and Santa Maria intrusions to to the east Donnelly et al.34 did not include the Jocotan- the south; Upper Cretaceous serpentinite allochthons north Chamelecon fault in their discussions of major strike-slip and south of the fault; middle-Miocene conglomerates north faults of Guatemala. A major difficulty confronting those of the fault with source rocks to the south; lead-zinc who propose 100s of km of offset on this fault is the deposits north and south of the fault9,11,25,53. In addition, observation that mapped traces of the fault in northwest major rivers and drainage divides have 130 km of left Honduras do not clearly connect with or project into the offset9. Displacement across the faults is believed to have Swan Fracture Zone, a requirement of a major transform begun about 10 Ma (middle Miocene). fault in this plate boundary zone. However, it can be argued In central and eastern Guatemala the Motagua fault that Neogene deformations in eastern Guatemala and north- zone consists of a wide zone of shears with two distinct, west Honduras that postdate its period of major strike-slip sub-parallel faults within a deep river valley. Left-lateral activity could have obscured a previous connection. Some displacement of between 0.7 to 3.4 m across the northern of strike-slip displacement is proven by the juxtaposition the two splays, the Cabanas fault, accompanied the 1976 across the Jocotan-Chamelecon fault zone of extended ter- Guatemala earthquake that had an epicentre in the Motagua rane against non-extended terrane (Figs 15.3, 15.4). How- Valley of eastern Guatemala75. The fault zone has very poor ever, major strike-slip displacement remains unproven but topographic expression in western Guatemala across the not out of the question. volcanic plateau, appearing on Landsat imagery as a diffuse zone of disconnected lineations. The fault traces may have been obscured by young volcanic ash, but absence of after- SUMMARY AND PROSPECTUS shocks following the 1976 event across that region suggests that Holocene displacement has been less than on the well- A bipartite metamorphic basement with Proterozoic and mapped faults to the east. Just west of the Mexico-Guate- mid-Palaeozoic ages of metamorphism has been reported mala frontier, where the fault zone emerges from the for the Guerrero block and northern part of the Maya block. volcanic rocks, a continuous, left-lateral strike-slip fault has The Chortis block also has two distinct metamorphic se- been mapped on trend with the Landsat lineations. This quences, but ages of these rocks are still uncertain. Contin- apparent westward extension has very little geomorphic ued work on ages of Chortis and Guerrero block expression, but has displaced Miocene intrusions by at least metamorphic rocks will be quite important in verifying the a few 10s of km. It is this author's view that the Motagua proposed match of the two. fault zone in westernmost Guatemala and Chiapas is an Carboniferous sedimentary sections of Chiapas, north- unlikely candidate for Neogene displacements of magnitude west Guatemala, and Belize could elucidate palaeo- greater than 100km. geographic settings and basin evolution during this period The map trace of the Jocotan-Chamelecon fault is that of proposed convergence betw een the Maya block and of a discontinuous arc comprised of 20 to 30 km segments North American plate. Especially important in this context offset by generally north-trending normal faults that bound are the volcanic flows and volcaniclastic sedimentary se- grabens, such as the Guatemala City graben (Fig. 15.3). The quences with radiolarian cherts within these sequences. fault zone is the boundary between uplifted basement ter - The Orizaba fault zone was the probable transform fault rane to the north and thick Neogene volcanic and volcani- at the western margin of the Gulf of Mexico during its clastic sedimentary rocks to the south. It is the southern opening in the late Jurassic. Throughout the remaining boundary of the ophiolite terrane of Central America and the Mesozoic the Orizaba fault zone was the western margin of northern boundary fault to a major Eocene Subinal Forma- a progressively deepening sedimentary basin across the tion basin (above)33. Donnelly et al.33, who first described Isthmus of Tehuantepec and other parts of southern Mexico the Jocotan fault, found that significant vertical separation and northern Guatemala. The Todos Santos-San Ricardo had occurred during the deposition of Aptian-Albian lime- siliciclastic-carbonate basins may have formed by transten-

280 Northern Central America

sion across individual faults within the system. Hints of 2Bateson, J. H. 1972. New interpretation of geology of Maya contemporaneous tectonism can be found in stratigraphic Mountains, British Honduras. American Association of literature on the great carbonate sequence, the Orizaba, Petroleum Geologists Bulletin, 56, 956-963. Sierra Madre, Ixcoy, and Coban-Campur Limestones, 3Bateson, J.H. & Hall, I.H.S. 1977. The geology of the Maya mostly reports of tectonic breccias in northern Guatemala7. It Mountains, Belize. Institute of Geological Sciences is the author's opinion that transtensional strike-slip faulting Overseas Memoir, 3, 43 pp. may prove to be the dominant tectonic process in the 4Bishop, W.F. 1980. Petroleum geology of northern Central evolution of Mesozoic and Cenozoic sedimentary basins America. Journal of Petroleum Geology, 3,3-59. across the Maya block. 5Blair, T.C. 1987a Tectonic and hydrologic control on Whether the Motagua Valley ophiolite terrane repre- cyclic alluvial fan, fluvial, and lacustrine rift-basin sents collision between the Maya and Chortis blocks, or sedimentation, Jurassic -lowermost Cretaceous Todos between the Maya block and an oceanic island arc, is a Santos Formation, Chiapas, Mexico. Journal of Sedi- problem that is still outstanding. mentary Petrology, 57, 845-862. Further documentation of late Cretaceous arc magma- 6Blair, T.C. 1987b. Mixed siliciclastic -caibonate marine tism in the Chiapas Massif and western Guatemala is re- and continental syn-rift sedimentation, Upper Jurassic - quired. The cross-section of tectonic elements from Gulf of lowermost Cretaceous Todos Santos and San Ricardo Tehuantepec across the Chiapas Massif (Fig. 15.8) is not Formations, western Chiapas, Mexico. Journal of Sedi- duplicated in eastern Guatemala. If the Soconuzco and Santa mentary Petrology, 58, 623-636. Maria granodiorites are offset 130 km across the Polochic 7Blount, D.N. & Moore, C.H. 1969. Depositional and non- fault, what has happened to the possible southeastward depositional carbonate breccias, Chiantla quadrangle, continuation of that arc across the other plate boundary Guatemala. Geological Society of America Bulletin, 80, faults? Intervening plutons in central and eastern Guatemala 429-442. and in the Sierra Omoa of Honduras of similar age and 8Burkart, B. 1978. Offset across the Polochic Fault of composition, lie along the northern marginal complex (see Guatamalaand Chiapas, Mexico. Geology, 6, 328-332. above) within the plate boundary zone. These may be dis - 9Burkart, B. 1983. Neogene North American-Caribbean placed segments of the same volcanic arc, lying now across plate boundary across northern Central America: offset inactive splays of the complex Maya-Chortis block trans - along the Polochic fault. Tectonophysics, 99, 251-270. form-fault boundary. The Soconuzco-Santa Maria plutons 10Burkart, B., demons, R.E. & Crane, D.C. 1973. Mesozoic and those of similar composition in Jamaica and the and Cenozoic stratigraphy of southeastern Guatemala Cayman Ridge are properly situated to conform to displaced American Association of Petroleum Geologists 71,73,81 terraces as suggested by the current models . Future Bulletin, 57, 63-73. studies of displacements across the Chortis-Maya block 11Burkart, B., Deaton, B.C., Dengo, C. & Moreno, G. 1987. plate boundary zone should investigate not only the intru- Tectonic wedges and offset Laramide structures along sions, but also other tectonic elements associated with the the Polochic fault of Guatemala and Chiapas, Mexico: arc. reaffirmation of large Neogene displacement. Tecton- There is the lingering problem that proven sinistral ics, 6, 411-422. displacement across faults of northern Central America is an 12Burkart, B. & Scotese, C. 1990. The Orizaba fault zone: order of magnitude less than the 1,100 km or more of link between the Mexican volcanic belt and strike-slip Cayman trough offset accepted by many woikers in the faults of northern Central America EOS, 71,1559., Caribbean. It appears difficult to reconcile 1,100-km of 13Burkart, B. & Self, S. 1985. Extension and rotation of displacement across northern Central America in the Neo- crustal blocks in northern Central America and effect gene. However, there are fewer constraints on major offset in on the volcanic arc. Geology, 13, 22-26. the Paleogene across the Jocotan-Chamelecon or Motagua 14Carpenter, R.H. 1954. Geology and ore deposits of the faults. Rosario mining district and San Juanciot Mountains, Honduras. Geological Society of America Bulletin, 65, 23-38. REFERENCES 15 Carr, M.J. & Stoiber, R.E. 1990. Volcanism: in Dengo, G. 1 & Case, J.E. (eds), The Geology of Nroth America. Anderson, T.H., Burkart, B., demons, R.E., Bohnenber- Volume H. The Caribbean Region, 375-391. Geologi- ger, O.K. & Blount, D.H. 1973. Geology of the cal Society of America, Boulder. western Altos Cuchumatanes, northwestern 16 Case, J.E. & Holcombe, T.L. 1980. Geologic -tectonic Guatemala. Geological Society of America Bulletin, map of the Caribbean region, scale 1:2,500,000. U.S. 84, 805-826.

281 BURKE BURKART

30 Geological Survey Miscellaneous Investigations Dengo, G. & Bohnenberger, O.K. 1967. Structural devel- Map 1-1100 opment of northern Central America. American Asso- 17Chubb, L.J. 1959. Upper Cretaceous of central Chiapas, ciation of Petroleum Geologists Memoir, 11,203-220. 31 Mexico. American Association of Petroleum Geolo- Dixon, C.G. 1956. Geology of southern British Honduras, gists Bulletin, 43, 725-756. with notes on adjacent areas. Government Printing 18Clemons, R.E., Anderson, T.H., Bohnenberger, O.H. & Office, Belize, 85 pp. 32 Buikart, B. 1974. Stratigraphic nomenclature of recog- Donnelly, T.W. 1989. Geologic history of the Caribbean nized Paleozoic and Mesozoic rocks of western and Central America: in Bally, A.W. & Palmer, A.R. Guatemala. American Association of Petroleum (eds), The Geology of North America. Volume A. An Geologists Bulletin, 58, 313-320. Overview. Geological Society of America, Boulder. 33 19Clemons, R.E. & Long, L.E. 1971. Petrologic and Rb-Sr DonneUy, T.W., Crane, D. & Burkart, B. 1968. Geologic isotopic study of the Chiquimula pluton, history of the landward extension of the Bartlett southeastern Guatemala. Geological Society of Trough-preliminary notes: in Saunders, J.B. (ed.), America Bulletin, 82, 2729-2740. Transactions of the Fourth Caribbean Geological Con- 20Cserna, Z. de. 197 la. Taconian (early Caledonian) defor- ference, Port-of-Spain, Trinidad, 28th March-12th mation in the Huasteca structural belt of eastern April, 1965, 225-228. Mexico. American Journal of Science, 271,544-550. 34Donnelly, T.W., Home, G.S., Finch, R.C. & Lopez Ra- 21Cserna, Z. de. 19715. Precambrian sedimentation, mos, E. 1990. Northern Central America; the Maya and tectonics and magmatism in Mexico. Geologischen Chortis blocks: in Dengo, G. & Case, J.E. (eds), The Rund-schau, 60, 1488-1513. Geology of North America. Volume H. The Caribbean 22Damon, P.E. & Montesinos, E. 1978. Late Cenozoic Region, 37-76. Geological Society of America, Boul- vol-canism and metallogenesis over an active Benioff der. zone in Chiapas, Mexico. Arizona Geological 35Engels, B. 1964. Geologische problematic und struk- Society Digest, 11,155-168. turanalyse Nikaraguas. Geologischen Rundschau, 54, 23Damon, P.E, Shafiqullah, M. & Claik, K.F. 1981. Age 758-795. trends of igneous activity in relation to metallogenesis 36Fakundiny, R.H. & Everett, J.R. 1976. Re-examination of in the southern Cordillera: in Dickinson, W.R. & Payne, the stratigraphy of the El Rosario and Comayagua W.D. (eds), Relations of Tectonics to ore deposits in quadrangles, Honduras. Publicaciones Geologicas del the southern Cordillera. Arizona Geological Society ICAITI, 5,31-42. Digest, 16,137-154. 37Finch, R.C. 1981. Mesozoic stratigraphy of central Hon- 24Dean, B.W. & Drake, C.L. 1978. Focal mechanism solu- duras. American Association of Petroleum Geologists tions and tectonics of the Middle America Arc. Bulletin, 65, 1320-1333. Journal of Geology, 86,111-128. 38Finch, R.C. & Ritchie, A.W. 1991. The Guayape fault 25Deaton, B.C. & Buikart, B. 1984. Time of sinistral slip of system, Honduras, Central America Journal of South the Polochic fault of Guatemala. Tectonophysics, 102, American Earth Sciences, 4, 43-60. 297-313. 39Fries, C. & Rincon-Qrta, C. 1965. Contribuciones del 26 Dengo, C.A. 1982. Structural analysis of the Polochic laboratorio de geocronometria, parte 2, Nuevas apor- fault zone in western Guatemala, Central America. taciones geocronologicas y tecnicas empleadas en el Unpublished Ph.D. thesis, College Station, Texas A & laboratorio de geocronometria. UniversidadAutonoma M University, 295 p. de Mexico, Instituto de Geologia, Bulletin, 73, 57-133. 27 Dengo, G. 1969. Problems of tectonic relations between 40Fries, C., Jr., Schmitter-Villeda, E., Damon, P.E. & Liv- Central America and the Caribbean. Gulf Coast ingston, D.E. 1962. Rocas precambricas de edad gren- Association of Geological Societies Transactions, viliana de la parte central de Oaxaca, en el sur de 19, 311-320. Mexico. Universidad Nacional Autonoma de Mexico, 28 Dengo, G. 1972. Review of Caribbean serpentinites and Instituto de Geologia, Bulletin, 64, 45-63. their tectonic implications. Geological Society of Amer- 41Frost, S.H & Langenheim, R.L. 1974, Cenozoic reef bio- ica Memoir, 132, 303-312. facies. Tertiary larger Foraminifera and 29 Dengo, G. 1985. Mid America; tectonic setting for the scleractinian corals from Chiapas, Mexico. Northern Pacific margin from southern Mexico to Illinois University Press, Dekalb, 388 p. northwestern Columbia: in Nairn, A.E.M. & Stehli, 42Gomberg, D.M., Banks, P.O. & McBirney, A.R 1968, F.G. (eds), The Ocean Basins and Margins. Volume Guatemala: Preliminary zircon ages from Central Cor- 7, 123-180. Ple num, New York. dillera. Science, 162, 121-122.

282 Northern Central America

43Gordon, M. B. 1990. Mesozoic igneous rocks on the sis, The University of Texas at Arlington, 196 pp. 56 Chortis block: implications for Caribbean reconstruc - Lopez Ramos, E. 1969. Marine Paleozoic rocks of tions. EOS, 70,1342. Mexico. American Association of Petroleum Geologists 44Guzman-Speziale, M., Pennington, W.D. & Matumoto, T. Bulletin, 52,2399-2417. 57 1981. The triple junction of the North America, Cocos, and Lopez Ramos, E. 1981. Geologia de Mexico, Volume III Caribbean plates: seismicity and tectonics. Tectonics, 8, (2nd edition). D.F., Mexico, 446 pp. 58 981-997. Malpica-Cruz. 1974. Estudio paleontologico y de micro- 45Hall, I.H.S & Bateson, J.H. 1972. Late Paleozoic lavas in facies del Paleozoico de la cuenca de Chicomuselo, Maya Mountains, British Honduras, and their possible Chiapas. Unpublished report, Institute Mexicano del regional significance. American Association of Petro- Petroleo. 9 leum Geologists Bulletin, 56, 950-956. Marcus, D.L. 1974. Igneous and metamorphic petrology of 46Fernandez, G.R. 1973. Paleogeografia del Paleozoico de Barillas quadrangle, northwestern Guatemala. Un- Chiapas. Boletin de la Asodacion Mexicana de Geologos published M.S. thesis, The University of Texas at Ar- Petroleros, 25, nos. 1-3, 77-134. lington, 117 pp. 47Hess, H. H. & Maxwell, J.C. 1953. Caribbean research 60McBimey, A.R. 1963. Geology of a part of the central project. Geological Society of America Bulletin, 64, 1-6. Guatemalan Cordillera. University of California Publi- 48Horne, G.S., Clark, G.S. & Pushkar, P. 1976. Pre-Creta- cations in Geological Sciences, 38,177-242. ceous rocks of northwestern Honduras: basement ter-rane 61McBirney, A.R. & Bass, M. 1969. Structural relations of in Sierra de Omoa. American Association of Petroleum pre-Mesozoic rocks of northern Central America. Geologists Bulletin, 60, 566-583. American Association of Petroleum Geologists Mem-oir, 49Horne, G.S., Finch, R.C. & Donnelly, T.W. 1990. The 11,269-280. Chortis Block: in Donnelly, T.W., Home, G.S., Finch, 62McBirney, A.R. & Williams, H. 1965. Volcanic geology of R.C. & Lopez Ramos, E., Northern Central America; the Nicaragua. University of California Publications in Maya and Chortis blocks: in Dengo, G. & Case, J.E. (eds), Geological Sciences, 55, 65 pp. The Geology of North America. Volume H. The 63Metal Mining Agency of Japan. 1978. Republic oj'Guatemala Caribbean Region, 55-72. Geological Society of America, report on geological survey of Cuchumatanes area, Boulder. western Guatemala, phase II. Unpublished report, Japan 50Horne, G.S., Pushkar, P. & Shafiqullah, M. 1976. Preliminary International Cooperation Agency, Government of Japan, K-Ar dates from the Laramide series of Central 43 pp. Honduras. Publicaciones Geologicas del ICAITI, 5, 91- 64Mills, R.A. & Hugh, K.E. 1974. Reconnaissance geologic map 98. of Mosquitia region, Honduras and Nicaragua Caribbean 51Kesler, S.E. 1971. Nature of ancestral erogenic zone in coasts. American Association of Petroleum Geologists nuclear Central America. American Association of Pe- Bulletin, 58, 189-207. troleum Geologists Bulletin, 55, 2116-2129. 65Mills, R. A, Hugh, K.E, Feray, D.E. & Swolfs, H.C. 1967. 52Kesler, S.E., Josey, W.L. & Collins, E.M. 1970. Basement Mesozoic stratigraphy of Honduras. American Association rocks of wes tern nuclear Central America: the Western of Petroleum Geologists Bulletin, 51, 1711-1780. Chuacus Group, Guatemala Geological Society of 66Molnar, P. & Sykes, L.R. 1969. Tectonics of the Carib bean America Bulletin, 81,3307-3322. and Middle America regions from focal mechanisms and 53Kesler, S.E., Levy, E. & Martin F., C. 1990. Metallogenic seismicity. Geological Society of America Bulletin, 80, evolution of the Caribbean region: in Dengo, G. & 1639-1684. Case, J.E. (eds), The Geology of North America. Volume 67Morovec, D. 1983. Study of the Concordia fault system H. The Caribbean Region, 459-482. Geological Society near Jerico, Chiapas, Mexico. Unpublished M.S. thesis, of America, Boulder. The University of Texas at Arlington, 155 pp. 54Lewis, J.F. & Draper, G. 1990. Geology and tectonic 68Mossman, W. & Viniegra-O., F. 1976. Complex fault evolution of the northern Caribbean Margin: in Dengo, G. structures in Veracruz province of Mexico. American & Case, J.E. (eds), The Geology of North America. Association ofPetroleum Geologists Bulletin, 60,379- Volume H. The Caribbean Region, 77-140. Geological 388. Society of America, Boulder. 69Muehlberger, W.R. & Ritchie, A.W. 1975. Caribbean- 55Litke, G.R. 1975. The stratigraphy and sedimentation of the American plate boundary in Guatemala and southern Barillas quadrangle, Department of Mexico as seen on Skylab IV orbital photography. Huehuetenango, Guatemala, CA. Unpublished M.S. the- Geology, 3,232-235. 70Peterson, J.A. 1983. Petroleum geology and resources of

283 BURKE BURKART

southeastern Mexico, northern Guatemala, and Belize. ternary faulting along the Caribbean-North American United States Geological Survey, Circular, 760, 44 pp. plate boundary in Central America. Tectonophysics, 52, 71Pindell, J.L. 1985. Alleghenian reconstruction and sub- 431-445. sequent evolution of the Gulf of Mexico, Bahamas and 87Steele, D.R. 1985. Physical stratigraphy and petrology of proto-Caribbean sea. Tectonics, 4, 1-39. the Cretaceous Sierra Madre Limestone, west-central 73Pindell, J.L. & Barrett, S.F. 1990. Geological evolution of Chiapas: in Contributions to the Stratigraphy of the the Caribbean region; a plate-tectonic perspective: in Sierra Madre Limestone (Cretaceous) of Chiapas, Part Dengo, G. & Case, J.E. (eds), The Geology of North 1. Universidad Nacional Autonoma de Mexico, Insti America. Volume H. The Caribbean Region, 405-432. tute de Geologia, Bulletin, 102, 1-101. Geological Society of America, Boulder. 88Thompson, M..L. & Miller, A.K. 1944. The Permian of 74Pindell, J. & Dewey, J.P. 1982. Permo-Triassic recon- southernmost Mexico and its fusulinid fauna. Journal struction of western Pangaea and the evolution of the of Paleontology, 18, 481-504. Gulf of Mexico/Caribbean region. Tectonics, 1, 179- 89Viniegra, O., F. 1971. Age and evolution of salt basins of 211. southeastern Mexico. American Association of Petro- 75Plafker, G. 1976. Tectonic aspects of the Guatemalan leum Geologists Bulletin, 55, 478-494. earthquake of 4 February 1976. Science, 193, 1201- 90Vinson, G.L. 1962. Upper Cretaceous and Tertiary strati- 1208. graphy of Guatemala. American Association of Petro- 76Ritchie, A.W. & Finch, R.C. 1985. Widespread Jurassic leum Geologists Bulletin, 17, 101-163. strata on the Choitis Block of the Caribbean Plate. 91Wadge, G. & Burke, K. 1983. Neogene Caribbean Plate Geological Society of America Abstracts with Pro- rotation and associated Central American tectonic evo- grams, 17, 700-701. lution. Tectonics, 2 ,633-643. 77Ritchie, A.W. & McDowell, F.W. 1979. K-Ar ages of 92Waite, E. 1985. Biostratigraphy and paleoenvironmental plutonic and volcanic rocks from the volcanic high- analysis of the Sierra Madre Limestone (Cretaceous), lands of Guatemala northwest of Guatemala City. Chiapas: in Contributions to the Stratigraphy of the Ischron/West, 25, 3-4. Sierra Madre Limestone (Cretaceous) of Chiapas, Part 78Roberts, RJ. & Irving, E.M. 1957. Mineral deposits of 2. Universidad Nacional Autonoma de Mexico, Insti Central America. U. S. Geological Survey Bulletin, tute de Geologia, Bulletin, 102,103-245. 1034,205pp. 93Webber, B.N. & Ojeda, J. 1957. Investigacion sobre laleri- 79Rosencrantz, E. & Sclater. J.G. 1986. Depth and age in the tas fosiles en las regiones sureste de Oaxaca y sur de Cayman Trough. Earth and Planetary Science Letters, Chiapas. Institute Nacional de Recursos Naturales, 79, 133-144. Mexico City, Boletin, 37,1-67. 80Rosenfeld, J.H. 1981. Geology of the western Sierra de 94Weber, H.S. 1979. On the lithology and stratigraphy of Santa Cruz, Guatemala, Central America: an ophiolite the "Estratos de Metapn" in the Republic of El Salva- sequence. Unpublished Ph.D thesis, State University of dor, Central America. Geologischen Jahrbuch, 37,31- New York at Binghamton, 313 pp. 54. 81Ross, M.I. & Scotese, C.R. 1988. A hierarchical tectonic 95Weyl, R. 1980. Geology of Central America. Gebriider model of the Gulf of Mexico and Caribbean region. Borntraeger, Berlin, 371 pp. Tectonophysics, 155,139-168. 96Williams, H. 1960. Volcanic history of the Guatemala 82Salvador, A. 1987. Late Triassic -Jurassic paleogeography Highlands. University of California Publications in and origin of the Gulf of Mexico. American Association Geological Sciences, 38, 1-86. of Petroleum Geologists Bulletin, 71, 419-451. 97Williams, H. & McBirney, A.R 1969. Volcanic history 83Sanchez-Barreda, L.A. 1981. Geologic evolution of the of Honduras. University of California Publications in continental margin of the GulfofTehuantepec in south- Geological Sciences, 85,1-101. eastern Mexico. Unpublished Ph.D. thesis, The 98Wilson, H.H. 1974. Crteaceous sedimentation and oro- University of Texas at Austin, 191 p. geny in nuclear Central America. American Associa- 84Sapper, K. 1899. Ueber Gebirgsbau und Boden des tion of Petroleum Geologists Bulletin, 58,1348-1396. nordlischen Mittelamerika Petermanns Mitteilung, 99Zoppis Bracci, L. 1957. Estudio geoldgico de la region de 27, no. 127,119pp. Palacagiiina y su deposito de antimonio. Boletin del 85 Sapper, K. 1937. Mittelamerika; Handbuch der Region- Service Geoldgico Nacional de Nicaragua, 1,29-34. alen Geologie. Steinmann & Wikkens, Heidelberg, 160 100Zoppis Bracci, L. & del Guidice, D. 1960. Reconoci- pp. miento geo!6gico del Valle de Punta Gorda. Boletin del 86 Schwartz, D.P., Cluff, L.S. & Donnelly, T.W. 1979. Qua- Servico Geologico Nacional de Nicaragua, 4, 61-83.

284 Index

Africa 16-17,20-21,24,26,28,231 Blanquilla province 249 African Plate see Africa Blue Mountains Block 111,113-114,117,121 A horizon 60 Bonaire 9,233,249-250,255-257,261 A'horizon 60 Bonaire Basin 9,15,249 A" horizon 4,48-49,53-54, 56,58-61 Brazil 230 Albian 17, 19-20, 24, 26-28, 57, 61, 71, 74, 80, 84, 112, Brazilian Shield 229 115,132,137,139,141,157,161,200,215,236,251, 255,275-277,280 calcareous nannofossils 115-116 Alps 3 Callovian 22,24,57,84 Altamiraterrane 132,144 Camaguey Block 67,73 amber 132 Cambrian 233,237,239,266,275 ammonites 57,76,133,200,210-211,213,249,251 Campanian 7, 17-20, 28-30, 49, 53, 57, 74, 76-81, 84, Andean basins 242-243 112-114, 116-117, 134, 137-139, 142, 144-145, 157, Andes 5-6,9-10,13,21-22,26,29,31,33-34,230,233,243 214-215,237,251,253,268,274,278 Anegada 152 Canada 267 Anegada Passage 5,8,151-152,154,167 Canouan 171-172 Anguilla 151,153,169 Carboniferous 6, 237, 239-240, 266-267, 271-273, 275- Antigua 151,153,169,171 278,280 Appalachians 3,80,267 Caribbean Mountains 9,193,209,233-237,241-243,261 Aptian 13, 16-19, 24, 26-28, 61, 72, 74, 84, 112, 115, Caribbean sea floor 41-64,139,144,229 132-133, 141, 145, 161, 211, 214, 270,273,275-276, Carriacou 171-173 279-280 Cayman Brae 87,89-90,93,98,100,103-104 Archaen 57,229 Cayman Islands 87-109 Aruba 9,57,233,241,249-252,257,261 Cayman Ridge 3,7-8,30,45,87,278-279,281 Atlantic 4,7,13,15-18,22,24,26,28,46,53,58,60-61, Cayman Trough (Trench) 3-5, 7-8,13,15,17-19, 31-32, 132,151-153,167-168,172,179,187,203 41,43,45,48,51,82,111,122,143,145,265,279,281 Aves Island 9,41 Cenomanian 17,81,84,132,134,157-158,161,200-204, Aves Ridge (Swell) 3, 5-6, 8-9, 17-19, 26, 30-31,41-42, 213,215,231,236-237,242,268,274,277 50-51,58,169 Cenozoic 6-10,13,17,30-34,43,45,54,59-60,67,77,115, 122, 140-141,144, 158,169,187,193,201-204,209, Bahamas, Bahamas Platform see Florida-Bahamas Platform 211,215,220,222,231,233,235,237,239-243,251, Baja Guajira basin 15,233,243 253,266,273-274,281 Bajocian 69,84 Central America 3-4,6-7,19,42,51,60-61,87,111,121, Barbados 8,31,46,169,179-192,240,261 144-145,243,265-284 Barbados Ridge 5,8-9,15,169,179,190 Central Cordillera see Cordillera Central Barbuda 151,153,169 Central Cuba Block 67,70-76 Barinas-Apure basin 233 Central Range (Trinidad) 209,211,213-215,223-224 Barremian 26-27,84,114-115,211,213-215 Cesar basin 15,233,243 Basse Terre (Guadeloupe) 171-173,175-176 Chiapas 19,21,265,267-268,270-274,276-281 Bathonian 22,84 Chiapas foldbelt see Sepur basin BeataRidge 3-5,41,43,48-49,55,57-59,133,143 Chicxulub Crater 6 Belize 15,45,121, 265-267,270,272-274,278,280 Choco Block 5-6,33,233,241-242 Bequia 171,173 Chorotega Block 5-6 Bermeja complex 19,32,156,161 Chortis Block 5-6, 18-19, 21, 24, 26, 28-31, 43, 53, 57, Berriasian 16,26,84 265-267,274-281 B horizon 53 Clarendon Block 111,114-117 B'horizon 53 Cocos Plate 6,33,265-267 B" horizon 4,29,43,48-49,53-59,139,142 Colombia 6,15,19,21,24,26-27,29,31,33-34,57,229, Blake Plateau 21 231-243

285 Colombian Basin 3-5, 7, 19, 41, 43, 45, 48-51, 53, 55, 233-234,237,239-240,243,251, 253,257,261,270, 57-58,60,64,111,139 274-275,277-278,280 Colombia Trench 241 Coniacian 4,27,72,77,84,116-117,133,137,139,142, Falcon Basin 15,32,233,235,243,261 231,237 Farallon Plate 17,26,28-29 Cordillera Central (Hispaniola) 129,133-134,140-142 Florida see Florida-Bahamas Platform Cordillera Central (SAm) 9-10,21,29,32,233,238,240- Florida-Bahamas Platform 3, 5, 7, 17,19, 21-22, 24, 26, 243 29-31,65,140-142,145,151,154,278 Cordillera de la Costa see Caribbean Mountains foraminiferans 57, 89, 91, 113-114, 120, 139, 160, 185, Cordillera de Merida (Venezuelan Andes) 9-10,233,235- 202-203,209-215,218,220,251,253,255,257 237,239,242 French Guyana 229-230 Cordillera Occidental 6,9,29,33,233,238,241-243 Cordillera Oriental (Hispaniola) 133 Galapagos 33 Cordillera Oriental(SAm) 9-10,33,233,237,239,242-243 Gondwana 16,21-22,26 Cordillera Septentrional 133,141 Grand Cayman 87,89-91,93,98-101,103-104 Costa Rica 6,15,19,28-29,43,57-58,61,277 Grande Terre (Guadeloupe) 169,174 Cretaceous 4, 6-9, 13, 17, 19-20, 22, 24, 26-31, 41-43, Gran Roque 258-259,261 50-51,53-59, 61,65, 67-71,73-75,77-78, 80-81, 84, Greater Antilles 5,7,18-19,26,28,30-31,50,65,111,129, 111-119, 121, 129-130, 132, 134-135, 137, 139-141, 142,145,151,154,156-161,167,278 143-145, 151, 154, 156-161, 168-169, 174, 193, Grenada 8,167,169-174,222,233 195-197,205, 209-215,217,222-224, 231, 233-243, Grenada Basin 3-5, 8-9, 13, 15, 30-32,41, 50, 167, 169, 249, 253,255,257,261,266,268,270-271,273- 172,174,243 276,278-281 Grenadines 8,167,169,171-172,174 Cretaceous/Tertiary boundary 6,139 Grenville Orogeny 7,80,267-268,270,277 Cuba 4-5,7-8,15,18,20,26,28-31,33,43,45,57,65-87, Guadeloupe 8,167,169,171,173-176 142,145,151 GuajiraPeninsula 20,24,233,241-243,261 Culebra 152,156-157 Guatemala 3,15,20,32,43,51,57,265-268,270-281 Curacao 9,57,61,233,249-251,253-255,257,261 Guerrero block 266-267,279-280 Curacao Ridge 9,242 Guinea Plateau 21 Gulf of Honduras 8,268,279 Daman 137,253,261 Gulf of Mexico 3, 5-7, 13-39, 4647, 80, 265, 273-274, Deep Sea Drilling Project see DSDP 278-280 Demarara Rise 58 Gulf of Tehuantepec 265-267,271,276,279 Desmoinsian 271 Gulf Stream 61 Devonian 237,239-240,270 Guyana 26,229-230 Dominica 171-174 Guyana Shield 34,229-231,233,237 Dominican Republic 5,15,26-27,34,129,134,137,139 DSDP 4, 24,41-43, 49, 51, 53-55, 57-59, 61, 139, 180, Hait i 5,7,51,53,58,121,129,134,137,139 182-183 Hanover Block 111,116-117 Duarte terrene 134,137,141-142,144 Hauterivian 84 Hess Escarpment 4,7,43 Eastern Cordillera see Cordillera Oriental Himalayas 3 Eastern Venezuela/Trinidad basin 5,18,20,32,231,233- Hispaniola 3-5, 7, 13, 17-19, 29, 32-34, 57, 61, 80-82, 235,237,243 129-150,152-154,159,275 Ecuador 26,28-29 Holocene 6-8, 17, 19-20, 28, 32-33, 45, 51, 80-82, 84, El Haul Swell 233 98-99,104,121-122,130,140,143-145,151,154-155, El Salvador 43,266,277 157, 159, 161, 168, 171, 182, 197, 235, 240, 249, Enriquillo basin 15,33 277-278,280 Eocene 4, 8,16-20,22,27,31-33,49,51, 58,60, 65,67, 72- Honduras 7,43,57,265-266,268,275-277,279-281 73, 77, 80-82, 84-85, 87, 113, 115-119, 121-122, Hotte-Selle-Bahorucoterrane 139,143-145 129,132-135,137,139-145, 151,154,156-157,159- hydrocarbons 19,115, 117, 154, 185,213,215,219-220, 161, 167-169,171-172, 174, 182-186, 202,212, 231, 223,231,233,235,273

286 inoceramids 113,116 Mesozoic 6,9,13,16,19,67,71,74,78,80-81,156-159, Made la Juventud (Isla de Pinos) Block 65,67,69,71,81 169, 195-205, 209-210, 214, 222-224, 231, 235- Isthmus of Tehuantepec 265,268,273-274,280 243, 258,261,266,268,270,273,275-281 Mexico 7,18-19,21-22,24,26,28,45,265-268,270,273, Jamaica 1,4-5,7,31,43,57,87,111-127,275,278,281 276-280 John Crow Mountains Belt 111,113,117,121 Mid-Atlantic Ridge 21,55 Jurassic 6-8,10,13,16-17,19,21-24,26-27,30-31,55-56, Middle America Trench 6,266-267,279 61,65,67-74,76-78, 80, 84,137,144,156,161,168, Middle Magdelana Basin 233,242-243 205,210,214,223,235-241,261,268,270,273,276, 278- Miocene 6,8,18,20,31-36,51,58,60,79,82,84,87,89, 280 91,93,100-101,113-114,117,120-122,129-130,132, 134-135, 137, 139-145, 151, 153-155, 157, 160- Kick 'em Jenny see Ronde-Kick 'em Jenny 161, 168-169, 171-172, 174, 182, 184-187, 195- Kimmeridgian 84,156 196, 203, 211-212, 215, 217-218, 223-224, 231, 233-235, 239-241,243,251,255,257,261,266,270- La Blanquilla 9,249-250,258-259 271,274,277, 280 La Desirade 13,19,26,169 MOHO 46,49-51,56,168,170 La Orchila 9,249-250,258-261 Mona Canyon 151-153 La Orchila Basin 9,249 Mona Passage 151,153-155 Laramide Orogeny 279 Montpelier-Newmarket Belt 111,121 LasAves 9,249-250,258 Montserrat 171,173-174,176 last interglacial see Sangamonian Mount Noroit 171 Leeward Antilles (ABC terrane) 249-258 MuertosTrough 4,41,133,135,139,151,153-154 Leonardian 271,278 Mustique 171,174 Les Saintes 171,173 Lesser Antilles 3-5,8-9,17-18,31-32,41,46,50-51,151- Nazca Plate 5,33,241 153,167-192,224 Neocomian 16,26,270,278 Limestone Caribees 8,168-169,174 Neogene 9,18-19,22,33,41,51,81-82,84,115,121,130, Little Cayman 87,89-90,93,103-104 134-135, 145, 181-183, 193, 200-203, 211, 223, Llanos Basins 230,233 233, 237-238,241-243,261,268,280-281 Loma Caribe/Tavera terrane 134,142 Netherlands Antilles 9,229,242-243,249-263 Los Frailes 9,250 Nevis 153,171,173 Los Hermanos 9,249-250,258-261 Nicaragua 7,43,265-266,268,275 Los Monjes 9,250 Nicaraguan Rise 3-5,7-8,15,30-31,41,43,48,55,57-58, Los Roques 9,249 60,111-113,121 Los Roques Basin 249 North America 4,6-8,13,15-22,24,26,29-31,33,61,65, Los Testigos 9,249-250,258 76, 80-81, 84, 87, 129-130, 144-145, 151, 154, 167- Lower Magdelana Basin 15 168,179,181,223,265-267,278-280 Luymes Bank 171,173 North American Plate see North America Northern Basin (Trinidad) 209,211-213,223-224 Maastrichtian 20,28-29,31,68,72-73,79-81,84,112-114, Northern Caribbean Plate Boundary Zone (NCPBZ) 4, 116,118,132, 137, 139, 144-145, 157-158,161,214- 31-33,81 215,236-237,251,257,274,279 Northern Range (Trinidad) 209-211, 214, 220, 223-224, Magdelana Fan 4,242 235 Maracaibo Basin 15-16, 18, 20, 22, 32, 233-234, 236, North Panama Deformed Belt 4-5 242-243 North Puerto Rican Basin 15,154-155 Maracaibo Block 33 Margarita 9,235,249-250,261 Oceanic allochthon (Barbados) 182-188,190 Marie Galante 169 ODP 80,182 Martinique 8,167-168,171-175 Offshore Drilling Project see ODP Maya Block see Yucatan/Maya Block Old Bahama Trench 133 Mayreau 171 Oligocene 7, 15, 19, 31-34, 67, 78, 82, 84, 87, 89, 100, Median Belt (Hispaniola) 134,137-138 134-135, 137, 139, 143-144, 151, 153-155, 157-158,

287

160-161,169, 171-172, 174, 184, 211, 217,231, 233- Quaternary 8, 18, 71, 84, 111, 120-122, 135, 169, 172, 235,239-241,243,251,253,268,277,279 188-190, 193, 195-197, 201-202, 220, 238,253, 255, Ontong-Java Plateau 58,61 257-258,261,267-268,277 Ordovician 233,235,237,239-240,266,270,277 radiolarians 19,49,57-58,60,73,76,132- Oriente Block 67,74-76,81-82 134,137,139,169,183,200,214,217,271,277,280 Oro terrane 133,141-142 Recent see Holocene ostracodes 117,185 Redonda 171,173 Oxfordian 24-25,69,76,84,273,278 Reykjanes Ridge 48 Rio San Juan/Puerto Plata/Pedro Garcia terrane132,144 Pacific 6,13,17-19,21, 24, 26-27,29,33, 53,58-61, 81, Ronde-Kick 'em Jenny 8,171-174 268,270,277,279 rudists 113,116-117,156,213,251,276 Palaeozoic 6-7, 9-10, 43, 231, 233, 235-243, 266, 268, 270-273,275-276,278,280 Saba 153,171,173-174 Paleocene 7, 9, 30, 32, 58-59, 67-68, 70, 77, 80-81, 84, 113-119, 121-122, 132, 134-135, 137, 139, 144-145, Saba Bank 9,15,151,153 151,169,185,212,233,236,240-241,243, 253, Saharan Shield 230-231 257,261,270,278 Samana terrane 130,132,142,144 Paleogene 6-9,18-20,24,26,29-31,33,41-42,55,57,67, Sangamonian 104,190,204,258 72-74, 78, 80-82, 84, 160, 181, 202, 205, 209, 211, San Jacinto terrane 233,242 213-214, 224, 233, 235, 237-239, 241-242, 249, San Jorge-Magdelana basin 233,243 274,281 San Juan Basin 15,17-18,32-33 Pan-African Orogeny 80 Santander Massif 239 Panama 3,6,15,28-29,33,43,57,233,242-243 Santonian 7,28-29,58-59,72, 84,115-117,133- Panama-Costa Rica Fans 4 135,137,144-145,211,214,242,251,253,270,273,27 Pangea 13,16,21-22 seamounts 53,137,141,171 Paraguana Peninsula 233,241-242 Seibo terrane 132-134,141-142,144 Pedro Bank 7 Senonian 251,255 Permian 6,16,233,237,239,266-267,271-273,277-279 Sepur basin 15,18,20,276 Peru 26,28 Sierra de Perija 9-10,233-234,237,239,242-243 Petit Canouan 171,173 Sierra Madre Occidental 19 petroleum see hydrocarbons Sierra Madre Oriental 22,26 Phanerozoic 57,60-61,230,237,239,241 Sierra Nevada de Santa Marta 9-10,19,233,241-243 Pinar del Rio Block 65,67-70 Silurian 239 Pleistocene 8,82,84,91-93,101,103,122,129,135,137, Sinu terrane 242 143-144, 169, 171, 174, 182, 185, 190, 197, 202- Sombrero 167 205,213,220,223,243,255,257 South America 3-4, 6,9-10, 13,15-22,24, 26,29-33,48, Pliocene 6,32,60,71,79,82,84,101,122,129,132,135, 50,59,61,76,80-81,84,167-169,179-181,183,209,223- 139-140, 144, 154, 160, 168-169, 171-172, 184- 224,229-247,261,278-279 185,195-197, 203-205, 211, 213, 220, 223, 231, South American Plate see South America 234,239,241,243,251,255,274 South Caribbean Deformed Belt 4-5,9,33,229,233,241- pollen (palynomorphs) 185,203,220 243,249 Precambrian 7,9-10,19,32,43,65,74,229,235-241,243, South Coast Basin (Puerto Rico) 154-155 267-268,270-271,275 Southern Basin (Trinidad) 209,215-220,223-224 Presqu'ile de Nord-Ouest Neiba terrane137,139,141,144 Southern Prism cover (Barbados) 182-185 Caribbean Plate Boundary Zone (SCPBZ) Proterozoic 229-231,267,280 32,220,229,233-237,239,243 pteropods 211 Southern Range (Trinidad) 209,215-220,223 Puerto Rico 3, 5, 7, 13, 17-19, 26, 32, 57, 60-61, St.Barthelemy 153,169,171, 174,253 145,151-165,275 St.Croix 151-155,158-161 Puerto Rico-northem Virgin Islands (PRNVT) Block St.Eustatius 171,173-174 St. John 152,157 151-154 St.Kitts 8,171-174 Puerto Rico Trench 3,34,45,151,153-154,159,167 St.Lucia 8,168,171,173-174,179

288 St. Martin 153,169,171 Union 171 St. Thomas 152,160 St. Vincent 8,171-176,233 USA 16,21-22,176 sub-Andean basins 233,236 Surname 229 Valanginian 24-25,84 Swan Island 87 Venezuela 9,15,17,24,27,31 -33,57,167-168,209,220,222- 223,229-243,249,253,261 Tertiary 6-9,19-20,29,32,51,67,70-71,77,87,111-113, Venezuelan Andes see Cordillera de Merida 115, 117-122, 129, 137, 139-140, 151, 154-161, Venezuelan Antilles 9,229,243,249-263 193, 195-196, 205, 220, 223,231, 233,236,238, Venezuelan Basin 3-6,9,19,41-43,45,48-51,53,55- 240-243,249-250,257-258,261, 268,271,274- 60,139,151-153 277,279 Venezuelan borderland 5,9,233,243 Tethys 19 Vieques 152,157 Tiero terrene 137,141-142,144-145 Virgilian 271 Tithonian 69,73,84,210,214 Virgin Gorda 152 Tobago 9,27,174,180,193-207,210,235,250,261 Virgin Islands 7,145,151-165 Tobago Trough 8,15,179-182,185 Virgin Islands Basin 151-152,154 Tortola 152 Volcanic Caribees 8,168-169,171-172,174 Tortue-Amina-Maimon terrane 134,141-142,144 Tortue Island 134 Wagwater Belt 111,117-118,121 Tortuga 250 Western Cordillera see Cordillera Occidental Trans-Amazonian Orogeny 229-230 Windward Passage 8 Triassic 6-7,13,15-16,19-24,233,236-237,239-240,267, Wolfcampian 271 278 Trinidad 1,16,20,24,57,176,180,209-229,233,235,250 YucatanBasin 3,5,8,13,15,30-31,41,45,47- Trinidad Province 209,211-220,223 51,55,60,81,265 Trois Riveres-Peralta terrane 137,141-142,144 Yucatan/Maya block/platform 3,5-6, 8,16-17,19,21- Turks and Caicos Islands 7 22,24,26,28-31,92,121,265-281

Turonian 4,29,49, 53, 58, 61,76,80, 84, 115, 117, 132, 137,139,157,161,215,237,249

289