See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/236589755

Evolution of the Southern Caribbean Plate Boundary

Article in Eos Transactions American Geophysical Union · January 2006 DOI: 10.1029/2006EO090001

CITATIONS READS 32 301

11 authors, including:

Alan Levander Michael Schmitz Rice University Fundación Venezolana de Investigaciones Sismológicas 366 PUBLICATIONS 5,552 CITATIONS 196 PUBLICATIONS 1,282 CITATIONS

SEE PROFILE SEE PROFILE

Dale S. Sawyer Marciane Magnani Rice University Universidade Federal da Paraíba 115 PUBLICATIONS 2,383 CITATIONS 121 PUBLICATIONS 987 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Seismic imaging of the lithospheric structure in intracratonic Illinois Basin area in central U.S. View project

Zircon Caribbean Project, Part 2: Giant PERLA Gas Field Basement and Paraguaná Peninsula, NW Venezuela View project

All content following this page was uploaded by Paul Mann on 28 May 2014.

The user has requested enhancement of the downloaded file. Eos, Vol. 87, No. 9, 28 February 2006

VOLUME 87 NUMBER 9 28 FEBRUARY 2006

EOS, TRANSACTIONS, AMERICAN GEOPHYSICAL UNION PAGES 97-104

tal landmass.The projects are designed to Evolution of the Southern investigate both the crust and upper mantle structure, as well as the evolution of the crust-mantle processes by which the Lee­ Caribbean Plate Boundary ward Antilles islands deform and accrete to South America (Figure 1). As part of the study of island arc accretion, the research PAGES 97,100 Lithospheric Investigation of Venezuela and the Antilles Arc Region; GEODINOS is the Vene­ will seek to determine how high-pressure/ It is generally accepted that the cores of zuelan project Recent Geodynamics Along the low-temperature (HP/LT) subduction-related the continents, called cratons, formed by the Northern Limit of South America. The two proj­ metamorphic rocks are exhumed.The proj­ accretion of island arcs into proto-conti- ects combined consist of a suite of geochemi- ect is also aiding earthquake hazard assess­ nents and then by proto-continental agglom­ cal, active and passive seismic, structural geol­ ment in the diffuse Southeast Caribbean eration to form the large continental masses. ogy sedimentary basin, and neotectonic plate boundary. Mantle-wedge processes, combined with studies involving about 60 scientists and stu­ higher melting temperatures during the dents from the United States and Venezuela. Tectonic Setting Archean (2.5-3.8 billion years ago) and pos­ One hypothesis that BOLIVAR and GEODI­ sibly thrust stacking of highly depleted NOS scientists are testing is that the Leeward The history of Caribbean tectonics and Archean oceanic lithosphere, produced a Antilles and related terrains are accreting to the Caribbean plate originates with the mis­ strong, buoyant, upper mantle chemical South America, contributing to the continen­ match in the opening of the North and boundary layer. This stabilizing mantle layer, known as the tectosphere, has shielded the BOLIVAR & GEODINOS Archean cratons from most subsequent tec­ tonic disruption and is highly depleted in 70 65 60 iron, providing the positive buoyancy that is 15 | 15 required to 'float' the continents more than four kilometers above the surrounding ocean basins. What is not clear is whether today the continental mass is growing, shrinking, or is at steady state. A number of continental growth curves have been proposed; the most widely accepted models call for rapid conti­ '% ky\Fa]c<>r> i J nental growth in the late Archean and Paleo- 10] 10 proterozoic (between 3.0 and -1.7 billion years ago), followed by slow growth to the present. Whether modern continental accre­ tion and something akin to tectosphere for­ mation are occurring today is an open ques­ tion. It is not clear how island arcs accrete to the continents, or if modern arcs contribute •UK Guyana Shield to continental growth. Seismic observations of arcs worldwide show that the crustal velocity structure is too fast, and hence the 5 chemical composition too silica-poor, to gen­ 70 60 erate an average continental crust without -6000 -1000 0 1000 8000 Bathymetrv \i\c\ a don (m) substantial chemical and/or mechanical refining during or subsequent to accretion. Fig. 1. Map of the southeastern Caribbean plate boundary, a diffuse plate boundary zone extend­ Two coordinated multi-disciplinary projects, ing from the southern Caribbean deformed belt in the north to the southern edge of the fore­ BOLIVAR and GEODINOS, are investigating land basin-fold and thrust belt system. The Leeward Antilles islands are A,Aruba; C, Curacao; continental growth and deformation processes B, Bonaire; and the Venezuelan archipelago is LA, Las Aves; LR, Los Roques; O, Orchilla; LB, La along the southeastern Caribbean-South Blanquilla; H, Los Hermanos; and T,Testigos. CAR UP is the Caribbean Large Igneous Province. American plate boundary (Figure 1). BOLIVAR Solid red lines on sea indicate BOUVAR marine reflection profiles; small white circles indicate is the U.S. project Broadband Onshore-Offshore coincident active source ocean-bottom seismographs (OBS); solid red lines on land indicate onshore-offshore and land refraction profiles, with red stars denoting land shots. Yellow triangles BY A. LEVANDER, M. SCHMITZ, H. G.AVE LALLEMANT, indicate the Venezuelan broadband seismograph network; white and blue trianges indicate the C. A. ZELT, D. S. SAWYER, M. B. MAGNANI, PMANN, broadband PASSCAL and Rice seismographs, respectively. White hexagons indicate broadband G. CHRISTESON, J.WRIGHT, G. PAVLIS, AND J. PINDELL OBSIP recorders. Original color image appears at the back of this volume. Eos, Vol. 87, No. 9, 28 February 2006

South Atlantic during the Late Jurassic-Early Cretaceous, between -125 and 180 million years ago. Westward migration of South America followed that of North America by -55 million years, with the result that today the South American Pacific coast is due south of the North American Atlantic coast. The basement rocks of most of the Carib­ bean islands are part of an island arc built in the Cretaceous along the western periph­ ery of the Americas known as the 'Great Arc of the Caribbean,' which was formed by eastward subduction of the now largely subducted Farallon plate [e.g.,Pindell and Dewey, 1982;Burke, 1988].At about 110 Ma, subduction polarity along the Great Arc flipped from east-dipping to west-dipping, initiating the eastward migration of the Caribbean, with the trench consuming a proto-Caribbean plate that had formed by rifting and drifting between North and South America. The Caribbean plate, thought to be once part of Farallon, has moved little in the hotspot reference frame; the Americas began moving westward past Distance from 12° North Latitude (km) it during the mid-Cretaceous, creating the Fig. 2. First arrival and PmP reflection travel-time tomography model along profile TRIN. The tectonic configuration that is seen today. profile extends 550 kilometers from the Venezuela basin across the Aves Ridge, the Lesser Antil­ Prior to 80 Ma, the modern Caribbean plate les Arc passes between Trinidad and Tobago, and ends in the Atlantic. Crustal thickness varies became an oceanic plateau by poorly from -27-28 kilometers beneath the ridges to <20 kilometers beneath the Grenada and Tobago understood processes that have been attrib­ basins. The velocity structure under the Aves Ridge and Lesser Antilles Arc is similar to that uted to fixed hot spots, transient plumes, beneath the Leeward Antilles (see Figure 3). Figure from G. Christeson, University of Texas Insti­ and ridge subduction. tute for Geophysics. Original color image appears at the back of this volume. Modern Caribbean tectonics began with the collision of the Caribbean plate with beneath or being overthrust by South Amer­ DINOS. At present, data from the main tran­ South America and the Bahamas bank at ica in the southwest. The Atlantic subduction sects are still being analyzed. about 55 Ma. Because the proto-Caribbean zone is the site of an active volcanic arc, To investigate upper mantle structure, a plate has been completely subducted, and whereas the southern Caribbean archipelago broadband seismic experiment was fielded much of the current Caribbean plate has is largely avolcanic. consisting of 27 PASSCAL, 8 Rice land instru­ been overprinted during the formation of ments, and 15 long-term deployment OBS the large igneous province, the only reliable Project Experiments instruments. Complementing these 50 instru­ magnetic anomaly data are in the Cayman ments was the 35-element, permanent broad­ trough spreading center, which also began Active seismic transects were acquired band seismograph network installed by the at -55 Ma. As a consequence, the pre-55 Ma during an expedition of the research vessel Venezuelan Foundation for Seismological history has been pieced together from (R/V) Maurice Ewing in April to June 2004 Investigations (FUNVISIS) shortly after 2000, global plate reconstructions and tectonic, along much of the plate boundary (Figure resulting in a combined network of more petrologic, and sedimentological studies 1). Five transects approximately along the than 80 broadband stations. [e.g.,Pindell et al, 1988]. 70th, 67th, 65th, and 64th meridians, and In addition to the geophysical experi­ Starting at 55 Ma, the mountain belts of from the Venezuela basin past Trinidad and ments are structural geology studies and northern South America were built diachron- Tobago to the Atlantic were heavily instru­ geothermometry and geobarometry analyses ously from west to east as the southern periph­ mented. The seismic images along each of along some of the principle transects and in ery of the Caribbean plate, fronted by the these transects provide a structural picture the Leeward Antilles. Age-dating using sensi­ Great Arc, collided obliquely with the South of the crust and mantle at different evolu­ tive high-resolution ion microprobe-reverse American margin [Pindell et al., 1988].This tionary stages of the Caribbean-South Ameri­ geometry (SHRIMP-RG) instrumentation is resulted in the time-transgressive development can plate interaction. This roughly gives providing a precise chronology of Leeward of the fold and thrust belt and foreland basin snapshots of the time-transgressive structural Antilles magmatism. Basin studies in both system, and exhumation of HP/LT rocks. Arc evolution of the margin from -55 to -15 Ma the onshore and offshore regions are provid­ volcanism along the southern edge of the on the meridional profiles, and the initial ing information on Caribbean tectonics and Great Arc in the collision zone was progres­ condition at 0 Ma along the profile past Trin­ paleogeography. sively shut off with the arc shedding its quies­ idad-Tobago (Profile TRIN, Figure 2). cent volcanic islands to the South American The meridional transects were acquired as Preliminary Results continental margin. Alternating transpression marine reflection profiles and onshore-off­ and transtension, (i.e., contraction and exten­ shore wide-angle seismic profiles, the latter The marine reflection/wide-angle profile sion superimposed on a largely transcurrent using as many as 49 ocean bottom seismo­ TRIN passing from the Atlantic to the Vene­ fault system) since the initial collision have graphs (OBS) and -550 land seismographs in zuela Basin provides a baseline seismic struc­ developed a strike-slip fault system rivaling Venezuela, with additional seismographs on ture of the Lesser Antilles Arc and Aves Ridge, California's San Andreas fault in length, com­ the Leeward Antilles. TRIN was acquired as a which are technically related to the islands plexity, and total displacement. reflection profile and OBS profile. Signals of the Leeward Antilles islands now accreting Today the Americas are subducting from the Ewing sound source array and eight to South America (Figures 1 and 2).Velocity beneath the Caribbean plate in the east, the on-land dynamite shots were recorded.Three models across the Leeward Antilles and con­ Cocos plate is subducting beneath it in the of the Venezuelan transects were later reac­ tinental mainland show large crustal thick­ west, and the Caribbean itself is subducting quired as low-fold reflection profiles for GEO­ ness variations: about 27 kilometers beneath Eos, Vol. 87, No. 9, 28 February 2006

Vp in km/s bean plate seaward of the plate boundary the R/V Ewing and R/V Seward Johnson II, for resulting from North America/South America their exceptional work during our field opera­ convergence. tions. Personnel from FUNVISIS and PDVSA/ Another striking observation is the associa­ INTEVEPand students from Simon Bolivar Uni­ tion of bodies of high compressional velocity versity the Central University of Venezuela, and a rocks (-7.0 kilometers per second) at shallow number of U.S. institutions, participated in the crustal depths with the Oca-San Sebastian-El field programs. We are grateful to the Nether­ Pilar strike-slip fault system, the major plate- lands Antilles Meteorological Service for help in boundary strike-slip faults.These rocks are the field. We thank the governments of Venezu­ interpreted as formerly subducted HP/LT ela, Aruba, and Curacao-Bonaire and Trinidad- metamorphic and/or mantle rocks being Tobago for permission to conduct our studies. exhumed by displacement partitioning and transtensional and transpressional processes References Fig. 3. P-uelocity profiles from the Lesser Antil­ which have occurred during 55 million years les, the Aves Ridge, and the Leeward Antilles of oblique plate convergence [e.g., Ave Lalle- Ave Lallemant, H. G. (1997),Transpression, displace­ (in red) compared with a number of island mant, 1997]. ment partitioning, and exhumation in the eastern arcs and the Sierra Nevada batholith, a former Caribbean/South American plate boundary zone, Initially a U.S.-Venezuelan collaboration, Tectonics, 16(2), 272-289. continental arc. Also shown is the Christensen the combined project has now expanded to Burke, K. (1988),Tectonic evolution of the Caribbean, and Mooney [1995] average crustal velocity include seismic installations on Aruba and Annu. Rev. Earth Planet. ScL, 16, 201-230. profile. The Caribbean arcs are substantially Curacao operated by the Netherland Antil­ Christensen, N. I., and W D. Mooney (1995), Seismic velocity structure and composition of the conti­ slower than Tonga and Bonin and somewhat les Meteorological Service, and cooperative slower than the Aleutians. They are somewhat nental crust: A global view.i Geophys. Res., 100(B7), seismic studies with the seismologists at the 9761-9788. higher in velocity than the Christensen-Mooney Seismic Research Unit of the University of Pindell,J.L.,and J.FDewey (1982),Permo-Triassic average, and substantially higher than the the West Indies, Trinidad-Tobago. Although reconstruction of western Pangea and the evolution Sierra Nevada. This indicates a composition of the Gulf of Mexico/Caribbean region, Tectonics, the initial objective of BOLIVAR/GEODINOS somewhat more mafic than the crustal aver­ 7(2), 179-211. age, although less mafic than other island arcs. was to study modern island arc accretion Pindell, J. L,S. Cande,W C. Pitman III, D. B. Rowley, J.FDewey J. LaBrecque, and W Haxby (1988), A Figure from A.Arogunmati, Rice University. and deformation as a continental growth process, an added benefit of the studies will plate-kinematic framework for models of Caribbean Original color image appears at the back of evolution, Tectonophysics, 155, 121-138. this volume. be to provide better locations and source White, R.V, J.Tarney A. C. Kerr, A. D. Saunders, PD. mechanisms of small magnitude earth­ Kempton, M. S. Pringle, and G.T. Klaver (1999), Modi­ the islands, less than 20 kilometers beneath quakes recorded by the various seismic fication of an oceanic plateau, Aruba, Dutch Carib­ arrays. This will better characterize the seis- bean: Implications for the generation of continental the basins, and -25-40 kilometers beneath crust, Lithos, 46(Y), 43-68. coastal Venezuela. Crustal velocities in the mogenic faults, seismo-tectonics, and earth­ Wright,J.E.,and S.J.Wyld (2004),Aruba and Curacao: Leeward and Lesser Antilles are similar to, but quake hazards of the entire region. Also, the Remnants of a collided Pacific oceanic plateau? lower than those of other modern island arcs large scale of the project (covering an area Initial geologic results from the BOLIVAR project, about the size of California and its conti­ Eos Trans. AGU, 55(47), Fall Meet. Suppl., Abstract (Figure 3).Though previous geologic and T33B-1367. geochemical investigations have called into nental margin) will provide a better tec­ question whether all of the Leeward Antilles tonic understanding of the evolution of the Author Information are island arc rocks in the traditional sense southeastern Caribbean plate boundary, an aid to hydrocarbon exploration in this [e.g., White et al, 1999; Wright and Wyld, Alan Levander, Department of Earth Science, energy-rich area. 2004], further investigations are still needed Rice University, Houston,Tex., E-mail: [email protected]; to draw definitive conclusions. Michael Schmitz, Venezuelan Foundation for Seis- Crustal thickening, thinning, and uplift are Acknowledgments mological Research, FUNVISIS, Caracas,Venezuela; observed throughout the plate boundary Hans G.Ave Lallemant, Colin A. Zelt, Dale S.Sawyer, zone, but particularly so in western Venezuela BOLIVAR was funded by grants from the U.S. and Maria B. Magnani, Department of Earth Science, in the Sierra de Falcon, where both the over- National Science Foundation's Continental Rice University; Paul Mann and Gail Christeson, Institute for Geophysics, University of Texas at Aus­ thrusting of Caribbean accretionary terranes Dynamics Program with additional funds from tin; James E.Wright, Department of Geology, Univer­ FUNVISIS and PDVSA/INTEVEP GEODINOS was onto Venezuela and underthrusting of the sity of Georgia, Athens; Gary L. Pavlis, Department of Caribbean plate beneath South America are funded by FONACIT grant G-2002000478.We Geological Sciences, Indiana University, Blooming- greatest. In eastern Venezuela, the seismic thank the personnel at the PASSCAL Instrument ton; and James Pindell, Department of Earth Sci­ data show incipient subduction of the Carib­ Facility and the OBSIP as well as the crews of ence, Rice University Eos, Vol. 87, No. 9, 28 February 2006

BOLIVAR & GEODINOS

-6000 -1000 0 1000 8000 Bathymetry/Elevation (m)

Fig. 1. Map of the southeastern Caribbean plate boundary, a diffuse plate boundary zone extend­ ing from the southern Caribbean deformed belt in the north to the southern edge of the fore­ land basin-fold and thrust belt system. The Leeward Antilles islands are A, Aruba; Q Curacao; B, Bonaire; and the Venezuelan archipelago is LA, Las Aves; LR, Los Roques; O, Orchilla; LB, La Blanquilla; H, Los Hermanos; and TTestigos. CAR LIP is the Caribbean Large Igneous Province. Solid red lines on sea indicate BOLIVAR marine reflection profiles; small white circles indicate coincident active source ocean-bottom seismographs (OBS); solid red lines on land indicate onshore-offshore and land refraction profiles, with red stars denoting land shots. Yellow triangles indicate the Venezuelan broadband seismograph network; white and blue trianges indicate the broadband PASSCAL and Rice seismographs, respectively. White hexagons indicate broadband OBSIP recorders.

Page 97

0° 10°E 20°E 30°E 40°E 50°E Fig. 1. The tracks of the MARE and ACSEX cruises. Black lines portray the surface ocean circula­ tion; red lines portray cruise tracks. Blue dots indicate the locations of current meter moorings; red dots indicate sediment traps. Cruise tracks indicate the investigations that were carried out on Agulhas rings in the southeastern Atlantic Ocean as well as the hydrographic observations undertaken of eddies in the Mozambique Channel and south of Madagascar. Eos, Vol. 87, No. 9, 28 February 2006

-200 -150 -100 -50 0 50 100 150 200 250 300 350 Distance from 12° North Latitude (km) Fig. 2. First arrival and PmP reflection travel-time tomography model along profile TRIN. The profile extends 550 kilometers from the Venezuela basin across the Aves Ridge, the Lesser Antil­ les Arc passes between Trinidad and Tobago, and ends in the Atlantic. Crustal thickness varies from -27-28 kilometers beneath the ridges to <20 kilometers beneath the Grenada and Tobago basins. The velocity structure under the Aves Ridge and Lesser Antilles Arc is similar to that beneath the Leeward Antilles (see Figure 3). Figure from G. Christeson, University of Texas Insti­ tute for Geophysics.

Page 100

Vp in km/s 5 6 7 8 0 SE Caribbean Arc ~—• Lesser Antilles Arc 5 »«. Aves Ridge » 67 W; ABC Rise 10 Other Arcs 15 Sierra Nevada Baihoiith ...... Aleutians 20 —— Benin Tonga 25 Christensen-Mooney 30 Average Continent

35

Fig. 3. P-velocity profiles from the Lesser Antil­ les, the Aves Ridge, and the Leeward Antilles (in red) compared with a number of island arcs and the Sierra Nevada batholith, a former continental arc. Also shown is the Christensen and Mooney [1995] average crustal velocity profile. The Caribbean arcs are substantially slower than Tonga and Bonin and somewhat slower than the Aleutians. They are somewhat higher in velocity than the Christensen-Mooney average, and substantially higher than the Sierra Nevada. This indicates a composition somewhat more mafic than the crustal aver­ age, although less mafic than other island arcs. Figure from A.Arogunmati, Rice University.

View publication stats