Journal of South American Earth Sciences 21 (2006) 383411 www.elsevier.

com/locate/jsames

Mesozoic transtensional basin history of the Eastern Cordillera, Colombian Andes: Inferences from tectonic models
L.F. Sarmiento-Rojas
b

a,*

, J.D. Van Wess b, S. Cloetingh

a Ecopetrol-Empresa Colombiana de Petroleos, P.O. Box 5938-6813-Bogota, Colombia Netherlands Institute of Applied Geoscience TNO National Geological Survey, Prins Hendriklaan 105 P.O. Box 80015 3508 TA Utretch, The Netherlands c Tectonics Group, Faculty of Earth Sciences, Free University, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands

Received 1 October 2004; accepted 1 March 2006

Abstract Backstripping analysis and forward modeling of 162 stratigraphic columns and wells of the Eastern Cordillera (EC), Llanos, and Magdalena Valley shows the Mesozoic Colombian Basin is marked by ve lithosphere stretching pulses. Three stretching events are suggested during the TriassicJurassic, but additional biostratigraphical data are needed to identify them precisely. The spatial distribution of lithosphere stretching values suggests that small, narrow (<150 km), asymmetric graben basins were located on opposite sides of the paleo-MagdalenaLa Salina fault system, which probably was active as a master transtensional or strike-slip fault system. Paleomagnetic data suggesting a signicant (at least 10) northward translation of terranes west of the Bucaramanga fault during the terrane with the Late Early Jurassic, and the similarity between the early Mesozoic stratigraphy and tectonic setting of the Payande Permian transtensional rift of the Eastern Cordillera of Peru and Bolivia indicate that the areas were adjacent in early Mesozoic times. New geochronological, petrological, stratigraphic, and structural research is necessary to test this hypothesis, including additional paleomagnetic investigations to determine the paleolatitudinal position of the Central Cordillera and adjacent tectonic terranes during the TriassicJurassic. Two stretching events are suggested for the Cretaceous: BerriasianHauterivian (144127 Ma) and Aptian Albian (121102 Ma). During the Early Cretaceous, marine facies accumulated on an extensional basin system. Shallow-marine sedimentation ended at the end of the Cretaceous due to the accretion of oceanic terranes of the Western Cordillera. In BerriasianHauterivian subsidence curves, isopach maps and paleomagnetic data imply a (>180 km) wide, asymmetrical, transtensional half-rift basin existed, divided by the Santander Floresta horst or high. The location of small mac intrusions coincides with areas of thin crust (crustal stretching factors >1.4) and maximum stretching of the subcrustal lithosphere. During the Aptianearly Albian, the basin extended toward the south in the Upper Magdalena Valley. Dierences between crustal and subcrustal stretching values suggest some lowermost crustal decoupling between the crust and subcrustal lithosphere or that increased thermal thinning aected the mantle lithosphere. Late Cretaceous subsidence was mainly driven by lithospheric cooling, water loading, and horizontal compressional stresses generated by collision of oceanic terranes in western Colombia. Triassic transtensional basins were narrow and increased in width during the Triassic and Jurassic. Cretaceous transtensional basins were wider than TriassicJurassic basins. During the Mesozoic, the strike-slip component gradually decreased at the expense of the increase of the extensional component, as suggested by paleomagnetic data and lithosphere stretching values. During the BerriasianHauterivian, the eastern side of the extensional basin may have ramo fault system. The western side probably developed by reactivation of an older Paleozoic rift system associated with the Guaica developed through reactivation of an earlier normal fault system developed during TriassicJurassic transtension. Alternatively, the eastern and western margins of the graben may have developed along older strike-slip faults, which were the boundaries of the accre ramo fault during the Late Triassic and Jurassic. The increasing width of the graben system likely tion of terranes west of the Guaica was the result of progressive tensional reactivation of preexisting upper crustal weakness zones. Lateral changes in Mesozoic sediment thickness suggest the reverse or thrust faults that now dene the eastern and western borders of the EC were originally normal faults ramo, La Salina, Bitu ima, with a strike-slip component that inverted during the Cenozoic Andean orogeny. Thus, the Guaica originally were transtensional faults. Their oblique orientation relative to the Mesozoic magmatic arc of Magdalena, and Boyaca

Corresponding author. Tel: +2345657. E-mail address: [email protected] (L.F. Sarmiento-Rojas).

0895-9811/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsames.2006.07.003

384

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

the Central Cordillera may be the result of oblique slip extension during the Cretaceous or inherited from the pre-Mesozoic structural grains. However, not all Mesozoic transtensional faults were inverted. 2006 Elsevier Ltd. All rights reserved.
Keywords: Rifting; Lithosphere stretching; Tectonic subsidence; Colombia; Mesozoic; Eastern Cordillera

1. Introduction This article focuses on the tectonic basin-forming processes of the Colombian Eastern Cordillera (EC, Fig. 1) during Mesozoic times, in terms of the geodynamic processes that govern deformation of the lithosphere. We compile local data into a regional geological model, analyze subsidence, and quantitatively model tectonic subsidence. We have studied tectonic subsidence signals that give important information about basin formation mechanisms. For this purpose, we analyze temporal and spatial basin subsidence patterns, quantitatively analyze tectonic subsidence, and forward model it to explain these patterns from within the framework of the geodynamic processes that formed the Mesozoic EC basin. In doing so, we address issues such as the relationship among basin development,

extensional episodes, plate-tectonic events, magmatic events, and basin geometry. Many features of these extensional basins and their underlying mechanisms are practically unknown. brard (1985) study the Fabre (1983a,b, 1987) and He subsidence of the eastern ank of the EC during the Cretaceous, identify the basin as produced by lithosphere extension, calculate tectonic subsidence curves, and, following the uniform instantaneous stretching model developed by McKenzie (1978), calculate lithosphere stretching factors close to 2. They distinguish an Early Cretaceous subsidence phase produced by rifting and Late Cretaceous subsidence produced by thermal decay after rifting. We assume several events of lithosphere stretching of nite duration in our study of tectonic subsidence and examine the possibility of dierentiating between crustal and subcrustal stretching

a de San Lucas; MA, Serran a de La Macarena. Eastern Cordillera regions: SM, Santander Massif; SF, SantanderFig. 1. Location map. SL, Serran Floresta high; MT, Magdalena Tablazo inverted subbasin; CO, Cocuy inverted subbasin; MF, Magdalena Valley foothills; LF, Llanos foothills; CU, n Massif; Romeral paleosuture in this map is the Cundinamarca inverted subbasin; SE, Southern Eastern Cordillera; QM, Quetame Massif; GM, Garzo western boundary of the Central cordillera and the Lower Magdalena Valley.

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

385

Fig. 2. Location of stratigraphic columns and wells and stratigraphic regional sections listed in Table 2. Numbers along sections refer to stratigraphic transect labels (for details, see Sarmiento, 2001).

that occurred in the Colombian Basin throughout the Mesozoic. An extensive data set of 162 stratigraphic columns and wells from the EC, Magdalena Valley (MV), and Llanos Orientales (LLA) areas (Fig. 2, see references in table 2.1 of Sarmiento, 2001) extracted from the literature, as well as data from Ecopetrol, are used. 2. Tectonic setting The EC is the eastern branch of the Colombian Andes (Fig. 1), which comprises three mountain ranges: the Eastern, Central, and Western Cordilleras, which merge southward into a single range. The MV separates the Eastern and Central Cordilleras, and the Cauca Valley separates the Central and Western Cordilleras. The EC and its bounding basins, the LLA in the east and MV in the west, dene the area studied herein. During the Mesozoic, the area of the EC was an extensional basin. During the Paleogene, according to some authors (e.g., Van der Hamen, 1961; Roeder and Chamber mez et al., 1999; Sarmiento, 2001), upthrulain, 1995; Go sted blocks and/or incipient inversion of the Mesozoic extensional basin may have occurred in the area of the EC. However, another view posits a unique simple foreland

basin existed, related to the topographic load of the Central Cordillera (e.g., Cooper et al., 1995). General consensus indicates that during the Neogene, the Mesozoic extensional basin became inverted, deformed, and uplifted to form the EC (Cooper et al., 1995). In the study area, during the Triassic and Jurassic, continental and volcanic facies were deposited in extensional basins (Mojica et al., 1996). During the Triassic, these basins seem related to Pangea rifting (Pindell and Dewey, 1982; Ross and Scotese, 1988; Cediel et al., 2003), and since the Jurassic, they developed behind a magmatic arc related to the subduction of the Pacic plates under the western border of South America (McCourt et al., 1984; Fabre, 1987; Toussaint and Restrepo, 1989; Cooper et al., 1995; Meschede and Frisch, 1998). During the Early Cretaceous, marine facies accumulated on a wide extensional basin system. Shallow-marine sedimentation ended at the end of the Cretaceous, due to the accretion of the oceanic terranes of the Western Cordillera. 2.1. Triassic and Jurassic plate tectonic interpretations The tectonic settings discussed subsequently assume that accretion of tectonic terranes east of the Romeral

386

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

paleosuture and west of the Llanos Orientales Basin (Fig. 1) occurred during the Paleozoic. Following this assumption for evolution in Colombia, the plate tectonic interpretations rely on two hypotheses. Although these interpretations dier in plate tectonic setting, an alternative hypothesis suggests these settings formed at dierent times. 1. Intracontinental rifting related to the breakup of Pangea (Pindell and Dewey, 1982; Ross and Scotese, 1988; Cediel et al., 2003) occurred during Triassic and Early Jurassic times. This hypothesis probably is more applicable to the northern part of Colombia and Venezuela and their separation from North America. 2. Backarc extension occurred behind a subduction-related magmatic arc (Maze, 1984; McCourt et al., 1984; Pindell and Erikson, 1993; Toussaint, 1995a,b; Pindell and Tabbutt, 1995; Meschede and Frisch, 1998). According to this hypothesis, the study area was located at the margin of the continent when active subduction of oceanic Pacic plates was occurring. This interpretation explains the Triassic and Early Jurassic extensional basins in the study area as backarc basins. However, a recent paleomagnetic investigation suggests that along-plate margin translations of terranes might have taken place during the early Mesozoic (Bayona et al., 2005), as inferred by Toussaint and Restrepo (1994). Bayona et al. (2005), on the basis of paleomagnetic data, suggest a signicant northward translation (at least 10) of terranes west of the Bucaramanga fault (Bucaramanga area, Floresta Massif and Upper Magdalena Valley terranes) with respect to the craton during the Early Jurassic and no signicant paleolatitude anomalies since then. Additional paleomagnetic data are needed to test and quantify the magnitude of translation of tectonic terranes (Bayona, personal communication, 2005). 2.2. Cretaceous plate tectonic interpretations For the Cretaceous, three alternative hypotheses suggest the processes that might have operated at dierent times: 1. Backarc extension (McCourt et al., 1984; Fabre, 1987; Toussaint and Restrepo, 1989; Cooper et al., 1995; Meschede and Frisch, 1998). Key evidence for this hypothesis is the existence of a subduction-related magmatic arc. 2. Passive margin (Pindell and Erikson, 1993; Pindell and Tabbutt, 1995). The scarcity of magmatic rocks in the basin seems to support this hypothesis, but a poorly dened Cretaceous magmatic arc (i.e., San Diego, Altavista, and Cambumbia stocks; Restrepo et al., 1991; Toussaint and Restrepo, 1994) in the Central Cordillera is dicult to explain. Alternatively, the Central Cordillera may have been located farther south of its present position during the Cretaceous.

3. Intracontinental rifting related to the opening of the Caribbean. Some authors (e.g., Geotec, 1992; Cediel et al., 2003) suggest a NW-SE graben developed in the northern part of the Central Cordillera during the Early Cretaceous. A poorly dened Cretaceous magmatic arc in the Central Cordillera is dicult to explain with this hypothesis. During the latest Cretaceous (post-Santonian), all plate tectonic interpretations propose a convergent margin west of Colombia. The Caribbean plate was moving northeastward relative to South America, while the Farallon plate was subducting west of southern Colombia (Pindell and Erikson, 1993; Pindell and Tabbutt, 1995). 3. Stratigraphy 3.1. Triassic and Jurassic synrift sedimentation The Triassic and Jurassic sedimentary record is present in several isolated outcrops (Fig. 3). Continental deposits with redbeds and volcanic eusive and pyroclastic deposits are dominant, though some marine facies appear locally. Triassic and Jurassic rocks were deposited in extensional basins mainly located in the Upper Magdalena Valley a de San Lucas, and the western ank of (UMV), Serran the EC (Mojica et al., 1996). Fig. 4 shows a stratigraphic synthesis modied after Mojica et al. (1996; Fig. 6). Triassic and Jurassic sedimentary rocks formed a sequence bounded by unconformities. The lower contact is marked by an unconformity, which is dominantly angular. The upper contact is dominantly unconformable but locally conformable. Jurassic deposits consisting of clastic facies deposited in dominantly continental environments are widely distributed. In those sections where there are some marine facies, they are under- and overlain by continental clastic facies. The ne-grained muddy marine facies record local marine incursions during the Late TriassicEarly Jurassic. Volcaniclastic, pyroclastic, and volcanic lavas are mainly restricted to the upper part of the Upper Triassic to the lower part of the Middle Jurassic (Mojica et al., 1996, Fig. 6). Facies and thickness similarities related to geographic positioning suggest that TriassicJurassic sedimentation occurred in two separate basin compartments, each with its own subsidence history and sedimentary ll (Figs. 3 and 4). 1. Upper Magdalena, Cienaga de Morrocoyal, and Sierra Nevada terrane (region A, Fig. 3). This section corresponds with the western part of the Chibcha terrane, , San as dened by Toussaint (1995a,b), or the Payande Lucas, and Sierra Nevada terranes, as proposed by Etayo-Serna et al. (1983). In this compartment, two marine incursions are recognizable: in the Late Triassic member of the Saldan (Norian?Rhetian, Chicala a

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

387

Fig. 3. Location of TriassicJurassic strata outcrops and stratigraphic sections. Labeling of stratigraphic sections according to Fig. 4. Stratigraphic section E represents the TriassicJurassic sedimentary record of the eastern part of the Chibcha terrane according to Toussaint (1995b). Stratigraphic section W , San Lucas, and Sierra Nevada represents the TriassicJurassic sedimentary record of the western part of the Chibcha terrane, equivalent to the Payande terranes of Etayo-Serna et al. (1983) (modied from Toussaint, 1995b). Inset: location of study area.

Formation, UMV) and in the Early Jurassic (Sinemurian, Morrocoyal, and Los Indios formations; ages according to Mojica et al., 1996). Continental sedimentation followed shallow-marine deposition in this compartment. Volcanic-related facies are volumetrically more important here than in the compartment formed by outcrops in the EC area (region B, Fig. 3). On the basis of paleomagnetic data from the UMV, Bayona et al. (2005) pro -San Lucas pose that the UMV and San Lucas (Payande terranes; Etayo-Serna et al., 1983) were located south of the equator before the Middle Jurassic and north of it during the late Jurassicmiddle Cretaceous, with a northward movement of at least 10 relative to the South American craton.

2. Eastern Cordillera (region B, Fig. 3). This area corresponds to the eastern side of the Chibcha terrane, as dened by Toussaint (1995a,b). Marine inuence on deposition is located within the Lower Jurassic (Montebel Formation; ages according to Mojica et al., 1996, Fig. 4). This compartment is characterized by an absence of Triassic-dated rocks and thick deposition of siliciclastic deposits on continental extensional basins (e.g., Gir n Formation). Bayona et al. (2005), on the basis of o paleomagnetic data from the Bucaramanga and Floresta areas, suggest part of this terrane located west of the Bucaramanga fault was located close to the equator during the early Middle Jurassic and moved northward about 4 relative to the stable craton. According to their

388

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 4. TriassicJurassic stratigraphic sections. Location of transects in Fig. 3. Vertical axis represents geological time, and horizontal axis represents present-day horizontal distance (km) without palinspastic restoration (modied from Mojica et al., 1996). Geyer (1982), however, suggests poorly n Formation) in the Cienaga de Morrocoyal area (north of Serran a de San Lucas) are Triassic and correlate with the Luisa fossiliferous facies (El Suda area. Formation of the Payande

reconstruction, during the Late TriassicMiddle Juras naga de Morrocoyal (Payande sic, the UMV and Cie San Lucas terranes) basin compartment was located south of the EC compartment.

3.2. Cretaceous sedimentation Most exposed rocks in the EC are Cretaceous in age. Fig. 5 shows a traverse time-stratigraphic cross-section of the basin. Cretaceous rocks, including locally the uppermost Jurassic and Paleocene deposits, form a megasequence bounded by regional unconformities that are at least locally angular. On a broad scale, Cretaceous rocks represent a major transgressiveregressive cycle with a maximum ooding surface close to the CenomanianTuronian boundary, corresponding to the maximum Mesozoic eustatic level (Fabre, 1985; Villamil, 1993; Fig. 5). Superimposed on this large-scale trend, several smaller transgressiveregressive cycles are present, suggesting an

oscillating relative tectonoeustatic level. Subsidence was rapid (Fabre, 1983a,b, 1987), but shallow-water sedimentation suggests deposition kept pace with it. The basin was a wide graben system oriented approximately NNE-SSW, divided into two subbasins (Tablazo and Cocuy, Fig. 6) by the Santander-Floresta paleo-Massif. To the north, these subbasins continued to the Machi rida Andes of Venezuela and ques trough in the Me a de Perija (Julivert, 1968; FabUribante trough in Serran re, 1985, 1987). To the south, these subbasins joined as the Cundinamarca subbasin (Bu rgl, 1961), where the thickness of the Cretaceous sections reaches a maximum (Figs. 2 and 5). Fabre (1987) and Sarmiento (1989) suggest the Cundinamarca trough was limited to the south and north by NW-SE transfer paleofaults (Fig. 6). The N-S lateral changes of thickness and an extensional relay system south of the Cundinamarca trough support the existence of NWSE-striking transfer faults. On the basis of the presence of Lower Cretaceous sedimentary rocks in the northern part of the Central Cordillera, Geotec (1992) suggests the existence of a NW-SE

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

389

Fig. 5. Cretaceous and Tertiary stratigraphic section. Location in Fig. 2. Vertical axis represents geological time according to the scale of Gradstein and Ogg (1996); horizontal axis represents present-day horizontal distance (km) without palinspastic restoration. This section has been constructed from earlier versions by Fabre (1985, 1987) and Cooper et al. (1995), modied according to sequence stratigraphy interpretations (Pimpirev et al., 1992; Fajardo n and Carrero, 1995; Villamil and Arango, 1998). et al., 1993; Villamil, 1993, 1994; Etayo-Serna, 1994; Ecopetrol, 1994; Rolo

graben, connected with the Cundinamarca subbasin. Compiled available thicknesses of these sediments for similar chronostratigraphic intervals, however, are signicantly thinner than those of the EC (Figs. 5, 7b, and 8a). If such a graben existed, in terms of subsidence, it was a minor feature compared with the grabens in the EC. These Lower Cretaceous marine sedimentary rocks in the Central Cordillera are strongly deformed and associated with mac volcanic rocks of ocean anity, which collectively have been named the Quebradagrande Complex by Nivia et al. mez (2005). These authors inter(1996) and Nivia and Go pret the rocks as originated within a marginal basin with oceanic crust formation developed along an older suture between dierent metamorphic terranes of the Central Cordillera (between the eastern Cajamarca Complex and the a Complex). Compiled thicknesses suggest western Arqu mezs a current N-S basin in agreement with Nivia and Go (2005) interpretation. 3.2.1. Early cretaceous synrift sedimentation Sedimentation started in the Tablazo subbasin in Jurassic times and continued during the Early Cretaceous locally, without a tectonic-related angular unconformity (e.g., Rio Lebrija section, Cediel, 1968). In other areas, Cretaceous sedimentary rocks rest with angular unconformity on earlier Mesozoic, Paleozoic, or even pre-Cambrian rocks. In the Tablazo subbasin, basal beds correspond to sandstones (Los Santos, Tambor, and Arcabuco formations) deposited in uvial environments, usually with detrital sediments derived from uplifted fault blocks

(Etayo-Serna and Laverde-Montan o, 1985). The amalgamation of channel beds indicates that the sediment supply exceeded the rate of basin subsidence at the JurassicCretaceous boundary. Bu rgl (1967) suggests that an initial marine incursion in the Cundinamarca subbasin ooded a continental area with a desert climate, which provided conditions for evaporite formation during the early stages of marine transgression. McLaughlin (1972) cites paleontological evidence of a BerriasianValanginian age for some evaporite occurrences. During the Berriasian, the sea ooded the basin from the northern part of the Central Cordillera toward the Cundinamarca subbasin (Etayo-Serna et al., 1976). Then, the sea advanced from the Cundinamarca subbasin to the north into two subbasins, while the Santander-Floresta paleo-Massif remained emergent (Etayo-Serna et al., 1976; Fabre, 1985, 1987; Sarmiento, 1989; Figs. 5, 7a, b, and 8a). The Tablazo and Cocuy subbasins started to form a single wide basin during the Hauterivian due to ooding of the Santander-Floresta paleohigh (Fabre, 1985) and base level rise. However, this intrabasinal high was a signicant barrier to sediment movement until the Aptian (Figs. 7a, b, and 8a) . To the south, both the Tablazo and Cocuy subbasins show a gradual increase in dark shale deposited in poorly oxygenated shallow-marine environments (Caqueza and Villeta groups; Fabre, 1985; Sarmiento, 1989). In the Cundinamarca subbasin, Cretaceous sedimentation started during the Tithonian?BerriasianValanginian with turbidite deposits in both the eastern (lower Caqueza Group;

390

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 6. Cretaceous basin compartments (subbasins) and their tectonic subsidence in meters. (A) Tectonic subsidence of the Tablazo and Cocuy subbasins. n; Mz, Manizales; I, Ibague ; N, Neiva; C, Cu cuta; Bu, Bucaramanga; T, Tunja; (B) Tectonic subsidence of the Cundinamarca subbasin. Cities: M, Medell ; V, Villavicencio; Y, Yopal; A, Arauca. B, Bogota

Pimpirev et al., 1992) and western (lower part of Utica sandstone, Murca Formation; Sarmiento, 1989; Moreno, 1990, 1991) anks (Figs. 7a, b, and 8a). Turbidite deposition prevailed until the Hauterivian at the eastern border of the basin (Caqueza Group, Pimpirev et al., 1992). During the earliest Cretaceous, basin subsidence exceeded sediment supply, resulting in retrogradation of the turbidite fan system, so distal fan sediments covered middle fan mouth channel deposits. Post-Berriasian, sediment supply increased and overwhelmed basin subsidence, resulting in progradation of the turbidite fan system (Pimpirev et al., 1992) and locally in progradation of deltaic sands during the Hauterivian (upper Utica sandstone; Sarmiento, 1989; Moreno, 1990). Dierential subsidence related to syn-sedimentary normal faulting caused unstable slopes on basin margins. These processes favored turbidite deposition during the early Cretaceous Aptian (lower Utica sandstone, Murca Formation, Sarmiento, 1989; Moreno, 1990, 1991; Socota Formation, a and Rodr guez, 1978; Caqueza Group, Pimpirev Polan et al., 1992; Figs. 7a, b, and 8a). An important transgression followed a relative sea-level rise during late Aptian time. The sea ooded all the area of the present EC, even south of the Cundinamarca subbasin

(Etayo-Serna et al., 1976; Etayo-Serna, 1994). Southward of the paleo-UVM, Cretaceous sedimentation started during the Aptian (Vergara and Pro ssl, 1994) in an extensional basin formed initially in Jurassic times. In the whole EC basin, abrupt lateral thickness changes and fault-related alluvial or turbidite deposition attest to local tectonic/dierential subsidence depositional conditions in the Early Cretaceous. Regional correlation of Early Cretaceous relative tectonoeustatic cycles is dicult to establish because of local active extensional tectonics. Since the Aptian, these relative tectonoeustatic cycles have become more tractable than the pre-Aptian cycles (Fig. 5). In the Central Cordillera, Lower Cretaceous sedimentary rocks (mudstones, chert, feldspathic sandstones, and conglomerates) of the Quebradagrande Complex are associated with an ophiolite suite of ultramac rocks, gabbros, basalts, breccias, and pyroclastic rocks usually aected by dynamic metamorphism (Moreno and Pardo, 2003). Depositional environments vary among uvial, coastal, deltaic, guez and Rojas, 1985). Evidence and marine shelf (Rodr indicates that during the middle Cretaceous, these rocks guez and Rojas, began compressive deformation (Rodr 1985), possibly as a result of the closure of the Quebradagrande marginal basin (Nivia et al., 1996).

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

391

Fig. 7. (a) BerriasianValanginian and (b) HauterivianBarremian thickness (m) without palinspastic restoration. Thick lines represent paleofaults believed to be active during BerriasianValanginian/HauterivianBarremian, respectively. (c) Contour map of crustal (d) lithosphere stretching factors calculated through forward modeling for the BerriasianHauterivian (144127 Ma, Cretaceous) stretching event without palinspastic restoration. Distribution of main Early Cretaceous faults and mac intrusions shown with circles: 1. Rio Nuevo diorite, 2. Rodrigoque micrograbbro, 3. Porritic basaltic lava, 4. Rio Cravo Sur microgabbro, 5. Pajarito, 6. Q. La Esperanza, 7. Q. Las Palomas, 8. Q. La Culebra, 9. Marl, 10. Q. Grande, 11. Q. La Chorrera, 12. La Chunchalita, 13. Q. La Fiebre, 14. Caceres, 15. La Corona, 16. Pacho, and 17. Rio Guacavia diorite. (d) Contour map of subcrustal (b) lithosphere stretching factors calculated through forward modeling for the BerriasianHauterivian (144127 Ma, Cretaceous) stretching event without n; Mz, Manizales; I, Ibague ; N, Neiva; C, Cu cuta; Bu, palinspastic restoration. Distribution of main Early Cretaceous faults is shown. Cities: M, Medell ; V, Villavicencio; Y, Yopal; A, Arauca. Bucaramanga; T, Tunja; B, Bogota

3.2.2. Cretaceous postrift sedimentation Several relative tectoeustatic level cycles have been proposed during the Late Cretaceous (Fig. 5). During the Albian, a relative base-level fall favored progradation of deltaic and littoral sands in the UMV (Caballos Formation, Etayo-Serna, 1994) and the eastern border of the basin (lower Une Formation, Fabre, 1985). During the middlelate Albian, the transition from near-shore

marine facies (e.g., Caballos, San Gil Inferior, and Socota formations) to outer shelf facies (e.g., Villeta Group, San formations) recorded a rise in Gil Superior, and Hilo relative tectonoeustatic levels (Villamil, 1993; Etayo-Serna, 1994). During late Albianearly Cenomanian, a relative tectonoeustatic level fall was recorded by progradation of the upper part of Une Formation and a generalized shallowing-upward facies trend (e.g., transition of San Gil

392

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 8. (a) Aptian thickness (m) without palinspastic restoration. Thick lines represent paleofaults believed to be active during Aptian time. (b) Contour map of crustal (d) lithosphere stretching factors calculated through forward modeling for the Aptian (121102.6 Ma, Cretaceous) stretching event without palinspastic restoration. Distribution of main Early Cretaceous faults is also shown. (c) Contour map of subcrustal (b) lithosphere stretching factors calculated through forward modeling for the Aptian (121102.6 Ma, Cretaceous) stretching event without palinspastic restoration. Distribution of main n; Mz, Manizales; I, Ibague ; N, Neiva; C, Cu ; V, cuta; Bu, Bucaramanga; T, Tunja; B, Bogota Early Cretaceous faults also shown. Cities: M, Medell Villavicencio; Y, Yopal; A, Arauca.

to unnamed shale; VillaSuperior to Churuvita and Hilo mil, 1993). During the late Cenomanian, Turonian, and Coniacian, the tectonoeustatic base level reached its maximum Mesozoic level. The sea ooded the entire northwestern corner of South America, and dark gray shale was deposited from Venezuela to northern Peru (Thery, 1982, in Fabre, 1985). In contrast to the EC, where the Cretaceous maximum ooding surface occurred at the CenomanianTuronian boundary, maximum ooding in the LLA occurred during

Formation; Fajthe Campanian (CS at the top of Gacheta ardo et al., 1993; Cooper et al., 1995, Fig. 4). Villamil (1993) recognized smaller relative tectonoeustatic level cycles during late Cenomanian, Turonian, and Coniacian times. A relative tectonoeustatic base level rise during the late Cenomanian (Villamil, 1993) induced a slight deepening of the basin and a notorious decrease of detrital supply to the basin (e.g., Frontera and lower part of San Rafael formations; Villamil, 1993). During the Turonian Coniacian, the present-day LLA foothills ooded (Cooper

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

393

et al., 1995) but not the entire LLA (Fig. 5). From the Middle Turonian to late Coniacian, a gradual progradation and shallowing upward (e.g., upper San Rafael Formation and Villeta Group in the upper MV) was related to a relative tectonoeustatic level fall (Villamil, 1993). In the UMV and during the late ConiacianSantonian, deepening of the basin and relative tectonoeustatic level rise occurred (transition from inner shelf uppermost Villeta Group to middle shelf lower chert unit of the Olini Group; Etayo-Serna, 1994, Fig. 2). During the Santonian, Campanian, Maastrichtian, and Paleocene, a general regression and progradation was recorded by littoral to transitional coastal plain facies (e.g., Guadalupe Group, Guaduas Formation). Guadalupe Group sands represent two cycles of westward shoreline progradation, aggradation, and retrogradation, dominated by high-energy, quartz-rich shoreface sandstones derived from the Guyana Shield (Cooper et al., 1995; Fig. 5). Regression did not occur continuously but with minor transgressive events recorded by ne-grained siliceous and phosphatic facies (Plaeners Formation, Olini Group, and upper La Luna Formation, Fo llmi et al., 1992; Fig. 5). At the end of Cretaceous (Maastrichtian), the gradual uplift of the western margin of the MV supplied clasts of metamorphic rocks accumulated by uvial systems close to the sea in a braided delta (Cimarrona Formation, mez and Pedraza, 1994). Go In the Central Cordillera east of the Quebradagrande Complex, no Late Cretaceous sedimentary record has been found. This area may have been uplifted during the latest Cretaceous and started to supply clastics to the east. 4. Methods 4.1. Subsidence analysis The stratigraphic record provides information about vertical crustal movements in a basin. Basin subsidence is the result of both a thermomechanical component called tectonic subsidence and a sediment and water loading component. Tectonic subsidence is undistorted basin subsidence in the absence of sedimentation and therefore is related to the geodynamics of the basin. To quantify the tectonic component of subsidence of the studied basin, we used the one-dimensional (1D) backstripping technique (Steckler and Watts, 1978; Bond and Kominz, 1984), as explained by Sclater and Christie (1980), Bond and Kominz (1984), and Bessis (1986). For this purpose, we calculate tectonic subsidence from the stratigraphic record, adopting local Airy isostasy to correct for the eect of sediment loading. The corrections for compaction follow porositydepth relationships on the basis of the observed lithologies using standard mean exponential relations and material parameters (cf. Sclater and Christie, 1980). Most stratigraphic columns are from published literature (see Sarmiento, 2001); well data are from Ecopetrol. We carefully checked the thickness of each stratigraphic

Table 1 Parameters used to calculate tectonic subsidence in the forward models Model parameters Initial lithospheric thickness Initial crustal thickness Asthenospheric temperature Thermal diusivity Surface crustal density Surface mantle density Seawater density Thermal expansion coecient Value 120 km 35 km 1333 C 1 106 m2 s1 2800 kg m3 3400 kg m3 1030 kg m3 3.2 105 C

column with available geological maps to avoid structural repetitions. We also veried the consistency of thickness between neighboring sections. The eects of paleobathymetry have been taken into account, using sedimentary facies and faunal content as interpreted in the literature. Ages are based on data in the literature, mainly the regional stratigraphic cross-sections presented by Cooper et al. (1995). To express ages in Ma, we used the geological time scale proposed by Gradstein and Ogg (1996). Unconformities are also included with ages and durations in Ma. Additional parameters for forward modeling are the initial crust and lithosphere thickness and densities. Table 1 provides the additional parameters we used in forward modeling, which are the average accepted values for normal continental lithosphere (Table 2). Pulses of fast tectonic subsidence of the basement have been interpreted in terms of tectonic processes. Only those pulses of fast tectonic subsidence correlatable to the normal movement component of fault activity are interpreted as produced by lithosphere extension or transtension, though normal faulting also may correspond to transtesional settings. Fast subsidence events related to crustal or lithosphere thinning also is common in transtensional basins or intra-arc basins (Ingersoll and Busby, 1995). In these settings, the extensional component is a major contributor to tectonic subsidence. The slower, generally later, rate of subsidence has been interpreted as produced by thermal reequilibration of the lithosphere following the thermal anomaly created by stretching. 4.2. Forward modeling of basin evolution To quantify the horizontal extensional movements responsible for the observed subsidence and establish a quantitative framework for the pulsating rift evolution of the lithosphere during Mesozoic basin formation, we quantify extension rates by forward modeling tectonic subsidence. We use an automated forward modeling technique (Van Wees et al., 1996b), which we explain briefly next. 4.2.1. Numerical model The forward modeling approach is based on lithospheric stretching assumptions (McKenzie, 1978; Royden and Keen, 1980). The extension factor d is used for crustal stretching and b for subcrustal stretching. For thermal

394

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Table 2 Cretaceous stretching events and stretching lithosphere factors from stratigraphic columns where the Cretaceous sedimentary record is present Basin compartmenta Stratigraphic columnb BerriasianHauterivian Event Start (Ma) End (Ma) Stretching factor b 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 5 5 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 9 9 9 9 9 9 72 Casabe-199 74 Cascajales-1 73 Infantas-1613 53 Lebrija-1 71 Llanito-1 70 Sabalo-1K 4 Arcabuco 8 Chima 10 Cimitarrra 24 Los Medios 25 Los Santos 47 Simacota 52 Tablazo 57 Vadorreal 58 Velez 59 Villa de Leiva 66 Chitasuga-1 22 La Calera 39 Quipile 48 Simijaca 75 Suba-2 76 Suesca-1 77 Suesca Norte-1 50 Sutamarchan 51 Tabio 60 Villeta 61 Yacop 3 Apulo 15 Fusagasuga 2 Alpujarra 12 Coello 16 Girardot 19 Guataqui 21 Itaibe 17 Melgar 29 Neiva 31 Ortega 35 Prado 36 Q Calambe 37 Q El Cobre 38 Q Olini 56 Chivata 64 Cormichoque-1 14 Floresta 18 Guaca 26 Matanza 54 Tibasosa 55 Tunja 78 Tunja-1 63 Bolivar-1 Corrales-1 6 Caqueza 28 Nazareth 34 Paz de Rio 44 Servita 33 Aguazul 9 Chita 7 Cocuy 20 Guateque 23 Labateca 69 Medina-1
c

Aptianearly Albian event Start (Ma) End (Ma) Stretching factor b 114 114 114 114 114 114 114.8 114.8 114.8 114.8 114.8 114.8 114.8 114.8 114.8 121 121 121 121 121 121 121 121 121 121 121 121 121 115.1 121 121 121 108.5 121 121 121 121 119.1 121 121 109.3 109.3 109.3 109.3 109.3 109.3 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 112.2 102.6 102.6 102.6 102.6 102.6 102.6 102.6 102.6 102.6 102.6 102.6 102.6 1.03 1.132 1.328 1 1.478 1.039 1.063 1.329 1.096 1.009 1.064 1 1.374 1.219 1.052 1.536 1.58 1.301 3.238 1.221 2.476 1.824 1.825 1.587 1.52 1.77 1.261 1.466 1.411 1.11 1.366 1.366 1.176 1.18 1.312 1.658 1.422 1.477 1.275 1.166 d 1.03 1.094 1.113 1 1.131 1.039 1.019 1.111 1.043 1.009 1.01 1 1.08 1.072 1.052 1.187 1.178 1.115 1.318 1.142 1.304 1.224 1.252 1.206 1.492 1.299 1.049 1.071 1.222 1.066 1.157 1.173 1.087 1.121 1.082 1.284 1.115 1.1 1.024 1.049

d 1.146 0.987 1.117 1.213 1.098 1.113 1.099 1.261 1.31 1.043 1.088 1.251 1.052 1.179 1.245 1.332 1.196 1.206 1.147 1.247 1.085 1.217 1.214 1.234 1.227 1.01 1.449 1.164 1.13

c c

c c

144 144 138 144 138 144 140 144 143 142 136.5 144 144 134.5 142 142 142 142 141 142 142 142 142 142 142 141 134.2 127.6 127.6

127.8 127.5 127.8 127 127.8 127.8 127.5 127.5 127.5 127.5 127.5 127.5 127.5 127.5 127.5 127.5 130 130 130 130 130 130 130 130 130 127.5 127.5 127 127

1.522 0.987 1.301 1.625 1.313 1.37 1.099 1.47 1.354 1.043 1.131 1.251 1.052 1.585 1.151 2.016 1.196 1.331 1.511 1.247 1.085 1.217 1.214 1.234 1.227 1.01 1.449 1.792 1.571

c c

c c

132.8 122.4 132.6 132 128.8 132.8 133.5 132 132 144 131.5 144 132 142 139 139 138 132 144

127 122.3 127 127 127 127 127 127 127 127.5 127.5 127 127.5 127.5 127.5 127.5 127.5 127.5 127.5

3.144 2.029 1.024 1.259 1.716 2.046 2.204 2.576 2.071 1.215 1.493 1.385 1.565 1.205 2.525 3.489 1.035 1.478 3.605

1.215 1.179 1.037 1.075 1.13 1.181 1.185 1.213 1.186 1.204 1.155 1.154 1.138 1.309 1.482 1.657 1.405 1.285 1.118

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411 Table 2 (continued) Basin compartmenta Stratigraphic columnb BerriasianHauterivian Event Start (Ma) End (Ma) Stretching factor b 9 9 9 9 9 10 27 32 40 46 49 67 Mojicones Pajarito R Cusay San Luis de Gaceno Sogamoso Cusiana-1X-2 139 142 136.5 144 136.5 127.5 127.5 127.5 127.5 127.5 1.36 1.169 1.643 1.076 1.459 d 1.23 1.303 1.35 1.352 1.354 112.3 112.2 1.095 Aptianearly Albian event Start (Ma) End (Ma) Stretching factor b d

395

1.025

Basin compartment numbers as shown in Fig. 6. Location of each stratigraphic column identied by number shown in Fig. 2. a Numbers indicate Cretaceous subbasins shown in Fig. 6. b Stratigraphic column number shown in Fig. 2. c Modeled using Triassic and Jurassic actual or inferred stratigraphy.

Fig. 9. Outline of the forward modeling technique. Explanation in the text (from Van Wees et al., 1996b).

calculations, we use a 1D numerical nite dierence model, which allows us to incorporate nite and multiple stretching phases. To handle the large number of wells and stretching phases in the forward model, we apply a numerical technique (Van Wees et al., 1996b), which automatically nds the best t stretching parameters for (part of) the subsidence data. In this procedure, we must specify the timing and duration of the rift phase, whereas the best t stretching values emerge by searching for the minimum of the mean square root F of the deviation in predicted and observed subsidence (Fig. 9), as a function of d and b, as follows: v uinum X 1 u 2 t F d; b 1 spp;i so;i ; num i1 where num is the number of subsidence data used in the tting procedure, and sp,i and so,i are predicted and observed subsidence values, respectively. For a rift phase, either uniform lithospheric stretching (d = b) (McKenzie, 1978) or

two-layered stretching (d b) (e.g., Royden and Keen, 1980) can be used. For uniform stretching, the solution to Eq. (1) requires that at least one observed subsidence data point is given after the onset of rifting, whereas twolayered stretching requires at least two data points. For polyphase stretching, the t is accomplished in sequential order. Initially, using a steady state thermal and compositional lithospheric conguration (cf. McKenzie, 1978), stretching parameters of the rst phase are determined by tting data points in the syn- and postrift time interval to the onset of the following phase. Subsequently, using the perturbed lithosphere conguration predicted at the onset of the second rift phase, stretching parameters are determined using subsidence data from its syn- and postrift time intervals to the next rifting phase. 4.2.2. Modeling procedure In the tting procedure, initial lithospheric conguration and thermal parameters are adopted as listed in Table 1. To t the data, we assume that each observed phase of

396

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

rapid tectonic subsidence should correspond to a stretching phase in the forward model. For these phases, we adopt a two-layered stretching model of the lithosphere (d b) to obtain the highest degree of freedom. However, we prefer to use a uniform stretching model (d = b) when uncertainty exists in estimating the stretching factors due to fewer data points or for relatively large age uncertainties, as is the case for the Triassic and Jurassic sedimentary record. Using the starting and nishing times previously determined for the stretching events, we calculate the lithosphere stretching factors that would produce theoretical subsidence curves similar to those observed. For the forward modeling, we include for most of the modeled locations the complete Mesozoic sedimentary section since the Triassic, even in those columns in which the pre-Mesozoic section is probably deep and does not crop out. In these cases, we use thicknesses interpolated from isopach maps. 5. Results In Fig. 10, we show the tectonic subsidence curves. The forward-modeled tectonic subsidence curves (Fig. 11) indicate a remarkably good t with the subsidence data. The lithosphere, crustal, and subcrustal stretching factors calculated for each stretching phase also are plotted in map view (Figs. 7, 8, and 12). 6. Tectonic subsidence and lithosphere stretching during Triassic and Jurassic time For the tectonic subsidence of the Triassic sedimentary record, we assume, following Geyer (1982), that the n Formation) of poorly fossiliferous sediments (El Suda a de the Cienaga de Morrocoyal north of the Serran San Lucas with lithology and relative stratigraphical position similar to the Luisa Formation of the Payande region are Triassic and time correlative and that Triassic sediments accumulated in the western ank of the EC. Using subsidence patterns, we attempt to dierentiate fast subsidence events. However, these attempts should be considered preliminary, because the continental Triassic and Jurassic poorly fossiliferous sedimentary record has relatively scarce and low-quality biostratigraphic data that make it dicult to establish clear time boundaries between the events. 6.1. Basin compartments Dierences in the shape of the subsidence curves in different areas conrm that TriassicJurassic sedimentation occurred in two separate basin compartments, each with its own subsidence history and sedimentary ll: -San Lucas terranes (UMV and Cie naga de 1. Payande Morrocoyal, region A, Fig. 3). 2. Eastern Cordillera (region B, Fig. 3).

6.2. TriassicJurassic subsidence and lithosphere stretching events 6.2.1. Early? Triassic event (variable in dierent columns, $248 to $235 Ma, time scale of Gradstein and Ogg, 1996) naga de This event is best represented in the UMVCie Morrocoyal with uniform stretching factors; b = d reach values of 1.23. Subsidence curves (Fig. 10), thickness variations, and the spatial distribution of stretching values (Fig. 12a) and paleomagnetic data (Bayona et al., 2005) suggest that small, narrow (150 km), transcurrent rift basins formed. Abrupt lateral changes of sediment thickness and facies in the UMV suggest dierential subsidence in dierent faulted blocks (Bayona et al., 1994; Mojica et al., 1996), which is common in transcurrent faults. Jurassic normal faults are described by Guillande (1988). The transtensional event correlates in time with magmatic arc activity in the Central Cordillera, interpreted by Aspden et al. (1987) as related to oblique subduction. The event also may have been related to intracontinental rifting (breakup of Pangea), as proposed elsewhere, particularly the separation between South and North America (e.g., Pindell and Dewey, 1982; Ross and Scotese, 1988; Cediel et al., 2003). However, considering that rifting and subsequent sea oor spreading, which originated the Gulf of Mexico, occurred during the Late Jurassic (Pessagno and Martin, 2003), this event probably relates to local transtensional events along the continental margin. 6.2.2. Latest TriassicMiddle Jurassic event ($208 to $185 Ma) Subsidence curves (Fig. 10), stratigraphic thickness, map distribution of stretching values, paleomagnetic data (Bayona et al., 2005), and fault distribution (Fig. 12b) suggest two narrow (<150 km), transtensional basins located at a de San Lucas (Fig. 1) the current location of Serran and the UMV. Fast subsidence favored marine ingression terrane). Probably in a dommainly in the south (Payande inant strike-slip setting, transtensional basins were separated by transpressional nodes, depending on the geometry of the strike-slip faults. The abundance of volcanic rocks in these transtensional basins implies a positive thermal anomaly that probably weakened the lithosphere. Lithosphere thermal doming could have produced the observed unconformities at the bottom of the synrift ll. Fast subsidence rates and a high level of volcanic activity are features commonly associated with oblique slip-rift zones (Ziegler, 1994). The width of these basins increased compared with the early Triassic (Mojica et al., 1996), suggesting the increasing width of the thermal weakened lithosphere. A weak lithosphere favored transtensional deformation. This extensional event correlates with magmatic arc activity in the Central Cordillera and may be interpreted as a backarc extension (Aspden et al., 1987). Rifting

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

397

and subsequent sea oor spreading, which originated the Gulf of Mexico, occurred during the Late Jurassic (Pessagno and Martin, 2003), so this event probably is related to local transtensional events along the continental margin.

6.2.3. Middle Jurassic event ($180 to $176 Ma) Paleogeography and stratigraphic thickness distributions indicate continued widening of transtensional basins, though they remained relatively narrow. Major depocenters developed in the MV and western ank of EC within

Fig. 10. Tectonic subsidence curves from the Mesozoic sedimentary record. Horizontal axis represents age in Ma. Vertical axis represents tectonic subsidence in meters obtained from backstripping analysis. Vertical shaded strips represent fast subsidence events. Numbers refer to basin compartments in Fig. 6. Note vertical bars represent the fast tectonic subsidence events.

398

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 10 (continued)

elongated NNE transcurrent grabens on opposite sides of the MagdalenaLa Salina fault system. The distribution of stretching b = d factor values (Fig. 12c) and paleomag-

netic data (Bayona et al., 2005) indicate rift basins located along the present-day western ank of the EC, with b = d values up to 1.39, and the paleo-MV. Moreover, large

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

399

Fig. 11. Forward modeled tectonic subsidence (continuous line) and observed tectonic subsidence (dots) curves (in m).

postrift Cretaceous subsidence along the western ank of can be explained only as therthe EC northwest of Bogota mal subsidence after a Jurassic stretching event. Therefore, these basins extended southward into the Cundinamarca region, possibly limited by the paleo-La Salina, paleoSuarez, and paleo-Boyaca fault systems. Small, isolated grabens developed in the Santander Massif (Geotec, (Shagam, 1975; Maze, 1992; Kammer, 1993), Perija rida Andes (Ricardi et al., 1990, in Mojica 1984), Me et al., 1996), LLA (Numpaque, 1986, in Cooper et al., 1995; Geotec, 1992), and Maracaibo (Shubert and Ricardi, 1980, in Mojica et al., 1996) areas. Volcanic activity rida decreased at this time, mainly occurring in the Me Andes (basalts in La Quinta Formation, Maze, 1984). These TriassicJurassic events correlate with magmatic arc activity in the Central Cordillera and Santander Massif. The calc-alkaline composition of these magmatic rocks suggests a magmatic arc related to high-angle subduction (Aspden et al., 1987). The most developed calc-alkaline magmatic -San Lucas terranes (i.e., plutonic bodies of the Payande and Segovia batholiths) tend to be adjacent to the Ibague west of these transtensional basins, which indicates these transtensional basins were located in a backarc setting close to the arc or an intraarc setting. Alternatively, these events may have been related to intracontinental rifting (breakup of Pangea), as proposed elsewhere, particularly the separation between South and North America (e.g., Pindell and Dewey, 1982; Ross and Scotese, 1988; Cediel et al., 2003).

However, considering that rifting and subsequent sea oor spreading, which originated the Gulf of Mexico, occurred during the Late Jurassic (Pessagno and Martin, 2003), these events probably relate to local transtensional events along the continental margin. Paleomagnetic data suggest that during the Late Triassic and Early Jurassic, the Santa Marta, Santander Massif, Floresta Massif, San Lucas, and Payande (UMV) magmatic arc segments aligned into a magmatic arc, striking parallel to the subduction zone (Bayona et al., 2005). This magmatic arc belt formed as a result of subduction along western Pangea, which resulted in the southward continuation of the early Mesozoic continental magmatic arc from the southwestern United States to Guatemala (Bayona et al., 2005). In the eastern part of the Chibcha terrane (EC east of the Bucaramanga fault), the main calc-alkaline plutonic bodies developed on the Santander Massif are positioned between the transtensional basins, suggesting an interarc setting. 7. Tectonic subsidence and lithosphere stretching during latest Jurassic and Cretaceous 7.1. Basin compartments Tectonic subsidence curves (Fig. 10) and restored thickness maps (Figs. 7a, b, and 8a) indicate several basin compartments (Fig. 6). In the northern part of the EC (Fig. 6), we recognize

400

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 12. (a) Contour map of lithosphere stretching factors (b = d) calculated through forward modeling for the Triassic (248.2235 Ma) stretching event, assuming there are Triassic sediments in the Cienaga de Morrocoyal area (Geyer, 1982) and that Triassic sediments accumulated in the western ank of the EC, without palinspastic restoration. Distribution of main early Mesozoic faults also shown. (b) Contour map of lithosphere stretching factors (b = d) calculated through forward modeling for the Early Jurassic (208185 Ma) stretching event without palinspastic restoration. Distribution of main early Mesozoic faults is shown. (c) Contour map of lithosphere stretching factors (b = d) calculated through forward modeling for the Jurassic (180.1176 Ma) , San Lucas terranes (Etayo-Serna stretching event without palinspastic restoration. Distribution of main Early Mesozoic faults also shown. (A) Payande et al., 1976), western part of Chibcha terrane (Toussaint, 1995a); (B) eastern part of Chibcha terrane (Toussaint, 1995a) and Guyana shield. Cities: M, n; Mz, Manizales; I, Ibague ; N, Neiva; C, Cu ; V, Villavicencio; Y, Yopal; A, Arauca. cuta; Bu, Bucaramanga; T, Tunja; B, Bogota Medell

1. Two subbasins: the Cocuy (region 9) and Tablazo (region 3) graben subbasins, separated by the less subsiding Santander-Floresta block (region 7), and 2. A regional westward decrease in tectonic subsidence with a maximum in the Cocuy subbasin (region 9) and a minimum in the Middle MV (region 2), which suggests a regional half-graben geometry for the whole basin.

In the southern part of the Cordillera at the latitude of , we recognize Bogota 1. A single extensional basin, the Cundinamarca subbasin (region 4, Fig. 6), and 2. BerriasianHauterivian tectonic subsidence, maximal in the eastern side of the Cundinamarca subbasin (region 8, Fig. 6), which indicates that a rst stretching event

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

401

ramo normal fault mainly aected the eastern Guaica system. During the Aptian, however, subsidence was maximal in the western side (region 4, Fig. 6), so a second stretching event mainly aected the western Bituima fault system. The total tectonic subsidence during the Cretaceous was slightly greater on the western side of the basin (region 4). In the western part on the Central Cordillera, a N-S (present day geometry) depocenter represents the Quebradagrande Complex marginal basin (Nivia et al., 1996; mez, 2005). Nivia and Go Tectonic subsidence in the UMV (region 6, Fig. 6), where marine sedimentation started in Aptian times, is signicantly less than that of the EC and Middle MV (region 2). In the easternmost LLA (region 10, Fig. 6), sedimentation started during the Late Cretaceous, and total tectonic subsidence during the Cretaceous was small compared with the EC and MV. Subsidence probably occurred through exural thermal subsidence (Watts et al., 1982) and water loading due to an increase in paleowater depth. 7.2. Latest JurassicEarly Cretaceous fast subsidence and lithosphere stretching events 7.2.1. Latest JurassicHauterivian event (variable in dierent stratigraphic columns, 144127 Ma) This event occurred in the area of the EC and is best represented in its eastern ank and west of the Bucaramanga fault. Subsidence curves (Fig. 10), thickness maps (Figs. 7a, b, and 8a), paleomagnetic data (Bayona et al., 2005), and the distribution of crustal d stretching factors (Fig. 7c) evidence a wide (>180 km), asymmetrical, transtensional half-graben basin divided by the Santander-Floresta high. Maximum tectonic subsidence and crustal stretching d up to 1.66 was associated with the pre-Guaica ramo normal master fault system, the eastern boundary of the graben (Figs. 7a and c). Strong changes in the thickness n Formation from 0 to more than 4 km across of the Giro ENE-WSW- and NE-SW-striking normal faults and local vertical axis rotations of fault-bounded blocks during this time, as documented by paleomagnetism, suggest an important control of normal faulting on deposition and basin conguration during the deposition of continental beds of n Formation (Bayona et al., 2005; Fig. 10, the Giro Tablazo-Lebrija). A second-order half-graben was located at the current location of the western ank of the EC with crustal stretching values up to 1.45 (Fig. 7c). This minor half-graben probably was associated with a paleonormal ma fault system, approximately following La Salina-Bitu fault system that was its western border. To the south, there was only one depocenter, limited to the south by a NW-SE vertical transfer fault. Early Cretaceous turbidites at both queza Group) of the extenanks (Murca Formation, Ca sional basin can be taken as evidence of tectonic instability associated with normal faulting movement. Branquet (1999) presents outcrop and seismic evidence of Cretaceous

normal faulting. Normal faults imaged on seismic sections (Fig. 13) conrm extensional tectonic movements that attest that this rapid subsidence event was produced by lithosphere stretching. As suggested by paleomagnetic data (Bayona et al., 2005), these faults may have had a dextral strike-slip component. Unlike the TriassicJurassic transtension, magmatic activity within the basin was reduced during the Early Cretaceous. Evidence of Early Cretaceous magmatism is limited to small, mac, igneous intrusions described by Fabre and Delaloye (1983) and Moreno and Concha (1993) and some volcanic input within Cretaceous shales (Rubiano, 1989; Villamil, 1994). Small mac intrusions described by Fabre and Delaloye (1983) coincide with areas of thin crust (crustal stretching factors >1.4) and places of maximum stretching of the subcrustal lithosphere (Fig. 7c). As a consequence of the depth-dependent lithosphere rheology assumed by the model, the results suggest that more intense stretching aected the subcrustal mantle lithosphere (Fig. 7d). Dierences between crustal and subcrustal stretching factors suggest some decoupling occurred between the crust and the subcrustal lithosphere or that increased thermal thinning aected the mantle lithosphere. The latter interpretation implies a considerable thermal anomaly produced by mantle lithosphere thinning, which seems supported by the presence of magmatic mac intrusions. During rifting, stress-induced lithosphere thinning causes adiabatic decompression of the lower lithosphere and asthenosphere, their partial melting, and the diapiric rise of melts into the zone of thinned lithosphere (Wilson, 1993). Although the 1D model cannot predict regional isostatic eects, the Lower Cretaceous unconformity on the rift margins (e.g., LLA) and locally on horst blocks (e.g., Santander-Floresta paleo-Massif) probably was produced by thermal uplift of rift shoulders, as suggested by the subcrustal stretching values. In addition, Jurassic magmatic activity in the Santander Massif contributes to relative thermal uplift and reduces tectonic subsidence. According to Ziegler (1994), unconformities on rift shoulders and intrabasinal fault blocks can be attributed to a footwall uplift in response to extensional unloading of the lithosphere. In general, the location of subcrustal and crustal stretched zones coincides, as a consequence of the 1D model assumption of local isostasy. However, where there is some oset, it indicates asymmetry in the basin, as supported by the general geometry of the basin. On the basis of subsidence analyses of Cretaceous stratigraphic columns brard (1985), using the of the EC, Fabre (1987) and He instantaneous stretching model of McKenzie (1978), calculate uniform b = d stretching factors up to 2 for the whole lithosphere. However, they lump the Cretaceous stretching events into a single instantaneous stretching event with an innite extension rate. The higher stretching values they obtain is a logical consequence of combining several stretching events with nite extension rates. According to the Cretaceous passive margin interpretation (Pindell and Erikson, 1993), active opening of the proto-Caribbean occurred north of Colombia and west of

402

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Fig. 13. Seismic sections in the Medina foothills area, along the eastern border of the Cundinamarca subbasin. Note normal fault evidence during the ramo paleofault system along the eastern border and contractional inversion of Cretaceous extensional faults during the Cretaceous (K) in the Guaica Paleogene evidenced by lateral changes of thickness of the Paleogene Carbonera and Mirador formations. Note thickness changes in the Cretaceous sedimentary ll (Linares, 1996). Location of seismic line is shown in Fig. 2.

the paleo-Central Cordillera. Moreno and Pardo (2003) correlate this latest JurassicEarly Cretaceous extensional phase, which originated the proto-Caribbean basin according to Pindell and Erikson (1993) reconstruction, with the Quebradagrande Complex marginal basin of Nivia et al. mez (2005). According to Moreno (1996) and Nivia and Go

and Pardo (2003), subduction of the western Quebradagrande proto-Caribbean crust beneath the Farallon plate originated a magmatic arc. To the east, a passive margin distant from the magmatic arc prevailed in the eastern side of the Quebradagrande Basin and east of the paleo-Central Cordillera (Moreno and Pardo, 2003, their Fig. 7). If such

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

403

an interpretation is valid for the BerriasianHauterivian, the event may be the result of stretching in the study area, which produced a failed-rifted arm related to a major opening of the proto-Caribbean oceanic basin. However, the presence of some Early Cretaceous plutonic rocks in the Central Cordillera east of the Quebradagrande Complex (i.e., lower Cre and Mariquita stocks, taceous K/Ar ages of the Samana Vesga and Barrero, 1978) is dicult to explain with a passive margin hypothesis. One possibility is that these ages are not reliable. An alternative plate tectonic reconstruction attributes Early Cretaceous extension of the paleo-EC to a backarc basin contemporaneous with reduced magmatic activity in the Central Cordillera. The proto-Caribbean eastern part of the Quebradagrande Complex basin crust may have subducted beneath the Central Cordillera to develop a magmatic arc. Although these alternative interpretations are controversial, the following evidence supports an EC basin located behind a partially emerged, less subsiding paleoCentral Cordillera (magmatic arc?): 1. Early Cretaceous igneous intrusions in the Central Cordillera east and west of the Quebradagrande Complex (e.g., San Diego, Cambumbia, and Mariquita stocks, Restrepo et al., 1991) dene one or two (?) magmatic arcs (produced by subduction of the W and E borders of the Queberadagrande basin crust?). However, such magmatic arcs are not well dened. 2. The presence in the western part of the Cundinamarca subbasin of Lower Cretaceous sandstones with abundant volcanic lithic fragments and feldspar derived from a western detrital source area, as indicated by paleocur tica sandstone; Sarrent data (Murca Formation and U miento, 1989; Moreno, 1990, 1991). 3. Volcanic lithic fragments in the mid-Cretaceous Caballos Formation of the UMV (Guerrero et al., 2000). 4. Progressive westerly onlap terminations of Cretaceous carbonates on the basement, observed in seismic lines in the western border of the Cesar Valley, northern Colombia (in Mesozoic times, part of the EC basinal area, Fig. 1; Audemard, 1991). 5. Stratigraphical and petrographical evidence suggesting that during the Berriasian (?)Valanginian, clastic sediments near San Felix in the western ank of the Central Cordillera (between Romeral and Palestina faults) came from erosion of uplifted areas with metamorphic rocks and small tectonic blocks with plutonic rocks (Rodr guez and Rojas, 1985). 6. Cretaceous volaniclastic rocks in the Central Cordillera guez and Rojas, 1985) consisting of mixtures of (Rodr pyroclastic and epiclastic fragments, probably derived from a magmatic arc. 7. Relatively high concentration of volcanogenic clay minerals in HauterivianBarremian (030%), middle Albian (021%), and Turonian (69%) shales of the Villeta Group (Rubiano, 1989) and the ValanginianHauterivian Rosablanca Formation (Moreno, 1989, in Rubiano, 1989) in the Cundinamarca subbasin. Thin beds of volca-

nogenic clays or bentonites in the CenomanianTuronian stratigraphic interval (Villamil and Arango, 1998) and Salada Member of the La Luna Formation (Patterson, 1970, in Rubiano, 1989), as well as subaqueous volcanic tus in La Frontera and La Luna formations in the MV (Restrepo-Pace, personal communication). 8. Jurassic (185 Ma) and Cretaceous (P77 Ma) zircon ssion-track ages from the Central Cordillera (Toro mez et al., 1999), evidence of uplift. In et al., 1999; Go the Ecuadorian Andes (Rivadeneira, 1996) and Central guez and Rojas, 1985), uplift or deforCordillera (Rodr mation have been suggested during the Late Cretaceous. 7.2.2. Aptianearly Albian event (121102 Ma) This fast subsidence event mainly occurred at the southern part of the western ank of the EC and the UMV, indicating asymmetry in the basin. In other localities, thermal subsidence is due to lithospheric cooling after prior lithospheric stretching events. During the BarremianAptian, the basin extended to the south in the UMV (Figs. 8a Memand 10). Turbiditic deposits of Aptian age (Socota a and Rodr guez, 1978) indicate tectonic instaber, Polan bility associated with the rapid subsidence event. The isopach map (Fig. 8a) suggests a master normal fault system, located approximately at the present-day Bituima Magdalena fault system, was active. In the UMV, a normal paleo-Chusma fault system probably also was active. Crustal stretching factors up to 1.4 are associated with the southern segment of the Bituima fault system and those up to 1.2 with the UMV (Fig. 8b). Because of the depth-dependent rheology assumed by the model, the results suggest that stretching aected the subcrustal mantle lithosphere more strongly. Subcrustal stretching values reach 3.24 at the southwestern ank of the EC and 1.6 at the UMV (Fig. 8c). Dierences between crustal and subcrustal stretching values suggest some decoupling between crust and subcrustal lithosphere or that an increased thermal thinning aected the mantle lithosphere. These results imply a thermal anomaly that probably is responsible for rift shoulder uplift, as evidenced from ssion-track data n Massif by Van der Wiel (1991) in the UMV and Garzo (Fig. 1). Van der Wiel (1991) interprets these ages as related to an orogenic event that aected the northwestern corner of South America between 100 and 80 Ma. However, this interpretation contradicts strong stratigraphic evidence of a subsiding basin at these localities. The zircon ages reect local uplift of faulted blocks located at rift margins, as demonstrated by ssion-track data from several rift basins (Van der Beek, 1995). Van der Beek (1995) explains rift margin uplift by mechanical support of rift anks, resulting from an upward state of exure. Evidence of magmatic activity in the basin is limited to some small mac intrusions (Fabre and Delaloye, 1983) and minor volcanic input (Rubiano, 1989; Villamil and Arango, 1998). Plate tectonic interpretations by Meschede and Frisch (1998) assume the beginning of subduction of the Farallon/Pacic plate under the Panama-Costa Rica arc west of

404

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

Colombia during the Aptianearly Albian. The alternative hypothesis of Pindell and Erikson (1993) assumes that during the Aptian, the proto-Caribbean lithosphere began to ime-Antilles arc, which was subduct under the Ama approaching the western margin of northern South America. 7.2.3. Cenomanian subsidence event (9893 Ma) This event occurred in the eastern ank of the EC. The Late CenomanianTuronian global sea level maximum correlates with it, suggesting that an increase in subsidence was driven by water load. However, the event aected only the eastern ank of the Cordillera, where the maximum Cretaceous thickness appears. At other localities, thermal subsidence resulted from lithospheric cooling after lithospheric stretching events. Although the maximum ooding surface for the Cretaceous sediments of the EC is the CenomanianTuronian boundary (Villamil and Arango, 1998), it is Campanian in the eastern LLA (Fajardo et al., 1993; Cooper et al., 1995; Fig. 5). If subduction of the Caribbeans thick and buoyant lithosphere (Duncan and Hargraves, 1984; Nivia, 1987; Hill, 1993; Meschede and Frisch, 1998) under South America was inhibited, it initiated uplift of the Central Cordillera, as evidenced by Late Cretaceous ssion-track mez et al., 1999; Toro ages from the Central Cordillera (Go et al., 1999) and exerted horizontal stresses on the northwestern margin of South America. Horizontal stresses can induce local exural lithosphere bending, which is maximal where the lithosphere is weakest (Cloetingh, 1988; Cloetingh and Kooi, 1992). This process probably enhanced the relative sea level rise, creating a maximum CenomanianTuronian marine ooding surface in the depocenter of the EC, characterized by weak lithosphere due to earlier stretching. In contrast, horizontal stress produced a submarine shallow water depth bulge in the LLA, which partially compensated for the maximum eustatic signal. 7.2.4. MaastrichtianPaleocene event (variable in dierent columns, 6854.8 Ma) This fast subsidence event aected the axial part of the EC, its eastern ank, and, locally, the westernmost part of the LLA. This event correlates in time with deformation and uplift in the Central Cordillera (Jaramillo, 1978, 1981; Cooper et al., 1995). All plate tectonic interpretations agree that during the latest Cretaceous and probably Paleocene, accretion of the Western Cordillera oceanic terranes along the Cauca-Patia fault occurred, producing deformation and uplift of the Central Cordillera. Increased subsidence in the ) axis of the Cundinamarca subbasin (Sabana de Bogota could result from increased horizontal compressional stress (Cloetingh, 1988; Cloetingh and Kooi, 1992) associated with the collision of the oceanic terranes of western Colombia and deformation and uplift of the Central Cordillera. Development of normal faults in the LLA area, as suggested by Kluth et al. (1997), could be the result of local tensional stresses in the developed exural bulge.

In conclusion, Jurassic (and probably post-Triassic) fast tectonic subsidence events are related to backarc to intra-arc transtensional lithospheric stretching, with a predominant dextral strike-slip component. Earlymiddle Cretaceous fast tectonic subsidence events seem related to backarc lithosphere stretching behind a poorly developed, subduction-related magmatic arc. An increase in distance from the backarc basin to the magmatic arc from the Jurassic to the Cretaceous has been observed elsewhere in basins that evolved from an initial intra-arc stage to a later backarc stage (Smith and Landis, 1995). Late Cretaceous subsidence was mainly thermal subsidence after the previous lithosphere stretching events, and localized discrete subsidence events probably result from lithosphere exural bending due to horizontal compressive stresses related to accretion of the oceanic terranes in western Colombia. 8. Discussion 8.1. Relationships between Mesozoic rifting and magmatism In the study area, abundant Upper TriassicLower Jurassic volcanic rocks are associated with moderate stretching factors (b = d up to 1.12). In contrast, the Cretaceous sedimentary record is almost devoid of volcanic rocks (only volcanic tus and minor mac intrusions) and associated with higher stretching factors (b up to 3, d up to 1.66). Thermal processes were more important than mechanical stretching during Late TriassicEarly Jurassic rifting than during Cretaceous extension. During the Late TriassicEarly Jurassic, abundant volcanic rocks suggest a positive thermal anomaly in the lithosphere but moderate lithosphere stretching. TriassicJurassic unconformities could have been produced by thermal uplift (active rifting?). Thermal doming results from progressive thinning of the higher density mantle lithosphere and its replacement by low-density asthenosphere (Bott, 1992). In contrast, during the Cretaceous, the less abundant volcanic rocks, absence of tectonically controlled unconformities, and large amount of tectonic subsidence suggest an absence of thermal doming. Small mac intrusions coincident with places of maximum crustal and mantle subcrustal stretching also suggest modest magmatism as a consequence of the extension of the lithosphere (passive rifting). The abundant Late TriassicEarly Jurassic volcanic rocks vary in composition from felsic to mac. Chemical analyses of La Quinta Formation volcanic rocks indicate a calc-alkaline composition in the AFM diagram and alkaline composition in the alkali-silica diagram (Toussaint, 1995b). Chemical analyses of the Saldan a Formation indicate a calc-alkaline composition, probably generated in a backarc environment (Bayona et al., 1994). The predominance of calc-alkaline compositions seems to suggest a convergent-related backarc extension rather than intracontinental rifting (Toussaint, 1995b).

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

405

Fig. 14. Event correlation between lithosphere stretching in the area of the EC and magmatic activity in the Central Cordillera. (1) Triassic, (2) Late TriassicEarly Jurassic, (3) Middle Jurassic, (4) Early Cretaceous BerriasianHauterivian, and (5) Aptian events. This correlation should be considered preliminary because original data are heterogeneous; 94% are K-Ar (biotite, hornblende, muscovite, or whole-rock), and 6% are Rb-Sr (hornblende/biotite or whole-rock). (a) Left panel: principal structural/plutonic zones of western Colombia. Right panel: age distribution of Mesozoic and Cenozoic plutonic activity in western Colombia (modied after Aspden et al., 1987). (b) Cumulative histogram of radiometric ages of plutonic bodies in Colombia (modied after Guillande, 1988). Periods of intense magmatic activity are characterized by rapid increase in the cumulative number of radiometric age determinations for a time interval (low slope).

8.2. Correlation between fast subsidence events and subduction-related magmatic arcs The inferred Mesozoic stretching events seem to correlate in time with reduced magmatic activity in the Central Cordillera (Fig. 14, modied from Aspden et al., 1987; Guillande, 1988). However, this preliminarily correlation should be tested with more and better data. If the calcalkaline (Alvarez, 1983) plutonic belts of the Central Cordillera and Santander Massif developed as subduction-related magmatic arcs during Mesozoic times, as suggested by Aspden et al. (1987), the extensional or transtensional basins behind them or close to them may be interpreted as backarc or intra-arc basins. Extensional

backarc basins develop when the velocity rollback, due to fast subduction, exceeds the oceanward convergence velocity of the overriding plate (Royden, 1993a,b). If magmatic arc activity decreases with the oceanward convergence velocity of the overriding plate, during times of reduced magmatic arc activity, a constant rollback velocity exceeds that velocity, increasing extension and subsidence in the backarc region. According to Aspden et al. (1987), the Triassic magmatic belt was controlled along strike-slip faults, as also supported by Restrepo-Pace (1995). The Jurassic magmatic arc may have been controlled by strike-slip faults, as suggested by paleomagnetic data (Bayona et al., 2005) and its elongated shape in map view, parallel to major strike-slip faults of major

406

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

and Segovia calc-alkaline plutonic bodies (i.e., Ibague batholiths). Along these faults, part of the magmatic belt might have moved northward, as suggested by paleomagnetic data (Bayona et al., 2005). Jurassic calc-alkaline plutonism and volcanism along the Central Cordillera and UMV is interpreted by Aspden et al. (1987) and Bayona et al. (1994) as a subduction-related magmatic arc, but Cretaceous plutonism developed sporadically only in the northern part of the Central Cordillera (Restrepo et al., 1991), whereas it is very extensive in Peru (Cobbing, 1982, in Aspden et al., 1987). Aspden et al. (1987) suggest oblique convergence and an oset in the subduction zone along a major NE-SW transform fault to account for the notable absence of Cretaceous plutonism in southern Colombia and Ecuador. 8.3. Geometry of transtensional basins 8.3.1. Triassic and Jurassic On the basis of paleomagnetic data, Bayona et al. (2005) terrane (UMV) was located south suggest that the Payande of equator (close to 10 S) before the Middle Jurassic; Sempere et al. (2002) recognize Late PermianMiddle Jurassic transtensional rifting in the EC of Peru and Bolivia (8 22 S). The early Mesozoic stratigraphy of the Peruvian Bolivian rift basin is very similar to that of the UMV terrane, sensu Etayo-Serna et al., 1983). In Peru (Payande and Bolivia, Late PermianTriassic, red to purple, coarsegrained, continental, synrift deposits (Mitu Group) resemble those of the Luisa Formation of the UMV; Late TriassicEarly Jurassic (Norian-Liassic), marine carbonate-dominated postrift deposits (Pucara Group) resemble Formation of the UMV; and Jurassic, those of the Payande brown-red mudstones and coarse-grained alluvial deposits (Sarayaquillo Formation) resemble the sedimentary component of the Saldan a Formation of the UMV. The similarity of the early Mesozoic stratigraphy in these areas initially was recognized by Geyer (1982). Synrift deposits of Peru and Bolivia commonly are interbedded with locally dominant volcanic and volcaniclastic rocks and/or intruded by subvolcanic to plutonic rocks (Sempere et al., 2002). A similar association with volcanic, volcaniclastic, and plutonic rocks is observed in the UMV. Paleomagnetic data (Bayona et al., 2005) and the similarity of the early Meso zoic stratigraphy and basin tectonic setting of the Payande terrane and PeruBolivia rift probably indicate that both areas were adjacent before the Middle Jurassic. In Peru and Bolivia, rifting events occurred during the Late PermianMiddle Jurassic interval (Sempere et al., 2002) and separated during the Late TriassicEarly Jurassic, thermal sag, postrift event during deposition of the Pucara Group (Sempere et al., 2002). Similar punctuated rift and lithosphere terrane). stretching events occurred in the UMV (Payande Geographically, however, between these Colombian and Peruvian/Bolivian transtensional basins in Ecuador during Jurassic, the accretion of terranes preserved as highly deformed metamorphic rocks has been proposed by Lither-

land et al. (1994). These Ecuadorian terranes were aected by intense ductile deformation and transcurrent faulting (Litherland et al., 1994). The early Mesozoic rift systems -San Lucas terranes in Colombia and the of the Payande Peruvian/Bolivian EC represent transtensional segments separated by a transpresional node in Ecuador, in line with the map view geometry of a major dextral strike-slip fault system. In Colombia, the calc-alkaline magmas likely intruded along active strike-slip faults and shear zones, developing a magmatic arc that may have resulted from a highly oblique subduction zone. Unlike the calc-alkaline magmatism of the Colombian TriassicJurassic magmatic arc segments, the Late PermianMiddle Jurassic rift system of the EC of Peru and Bolivia is characterized by intense magmatism, predominantly alkaline to tholeiitic, probably related to intraplate magmatism and lithospheric thinning (Sempere et al., 2002). In southern Peru and Bolivia, the Late Permian Middle Jurassic rift system is located behind the Jurassic magmatic arc (Sempere et al., 2002, Fig. 3). In central Peru, plutons probably were emplaced in the rift roots (Sempere -San Lucas terranes et al., 2002). The Colombian Payande may have originated as fragments of thinned lithosphere separated from the PeruBolivia rift system by intraplate rifting, which later moved north along strike-slip faults. The Late PermianMiddle Jurassic rift system of the EC of Peru and Bolivia is interpreted by Sempere et al. (2002) as a transcurrent rift system in which transtensional segments were separated by transpressional nodes. This transcurrent setting also could explain Triassic plutons in Peru and Bolivia with deformation contemporaneous to their emplacement (Sempere et al., 2002), as well as thenfoliated Triassic plutons described in Colombia (Aspden et al., 1987). However, new paleomagnetic, geochronological, isotopic, petrological, structural, and stratigraphic research in necessary to test this hypothesis. 8.3.2. Cretaceous During the Cretaceous, the Colombian transtensional basins increased in width. During the Mesozoic, the strike-slip component gradually decreased at the expense of increases in the extensional component, as preliminarily suggested by paleomagnetic data (Bayona et al., 2005) and lithosphere stretching values. As in other areas (e.g., East African rift), probably during the initial Cretaceous extensional stages, reactivation of crustal discontinuities led to the subsidence of isolated grabens linked by shear zones. The increase in width probably was the result of increasing strain, with grabens propagating toward one another, coalescing, and evolving into a generally continuous rift system (Ziegler, 1994). Whether the Colombian Mesozoic extensional basins were pure shear or simple shear extensional basins is dicult to demonstrate. Probably both mechanisms operated; these extensional basin models should be viewed as end-member cases. If the orientation of preexisting crustal discontinuities is such that they could not be reactivated by the stress system governing the

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

407

evolution of the transtensional basin, new faults would develop, and pure shear deformation will likely prevail (Ziegler, 1994). This mechanism may be applicable to the TriassicJurassic extensional basin system in Colombia. However, if the upper crust is weakened by the presence of preexisting crustal discontinuities and is oriented to favor reactivation under the prevailing tensional stress eld, they will present zones of preferential strain concentration, which can result in simple shear deformation (Ziegler, 1994). This mechanism may explain the development of the paleo-EC rift system during the Early Cretaceous: The eastern side of the extensional basin developed during the BerriasianHauterivian by reactivation of an older Paleozo ramo paleofault (Hossack ic rift system along the Guaica et al., 1999), and the western side developed by reactivation of earlier transtensional fault systems from the Triassic Jurassic. Another alternative is that the eastern and western margins of the graben developed along older strike-slip fault systems, which bounded the accretion of tectonic terr ramo paleofault system during the anes west of the Guaica Late Triassic and Jurassic (Bayona et al., 2005). The rheological properties of the lithosphere control the depth at which tensional necking occurs and whether a rift zone is exed up- or downward (Ziegler, 1994). A deep lithosphere necking level causes upward exure of the rift zone, whereas necking at shallow crustal levels causes downward exure and the absence of shoulder uplifts (Kooi et al., 1992; Ziegler, 1994). Similar deep levels of necking in the eastern side of the Early Cretaceous extensional basin system may have generated shoulder uplift in the LLA area during the Early Cretaceous. Coarse detrital fragments in the Lower Cretaceous Brechas de Buenavista (Pimpirev et al., 1992) and Calizas del Guavio (Conglomeguez, 1976) formations rado de Miralindo, Ulloa and Rodr could be derived from this graben shoulder. In contrast, in the western margin of the Early Cretaceous extensional basin, sedimentation was more continuous from Jurassic to Early Cretaceous times, which implies the downward exure of the rift shoulders and thus a shallower level of necking during Early Cretaceous times. On a lithospheric scale, the location of rift systems is controlled by the location of weakness zones in the lithosphere, which in turn depends on its thermal state and crustal thickness. At crustal scales, the composition, thickness of its mechanically strong upper layer, and availability of internal discontinuities (which can tensionally be reactivated) are also important controls for rift location (Ziegler, 1994). The overall present-day pattern of these transtensional basins for the Cretaceous indicates several grabens oriented chelon pattern, in contrast with the NNE-SSW in an en e more N-Soriented Central Cordillera (e.g., Mojica et al., 1996). Some authors (Fabre, 1987; Sarmiento, 1989; Geotec, 1992; Mojica et al., 1996) suggest that some NW-SE faults probably represented transfer faults. Features such as the Nazareth NW-SE fault (Fig. 7c), which limits the Early Cretaceous basin to the south (Fabre, 1987), or the NW-SE alignment connecting the two emerald districts of the EC

(Fig. 7c, Sarmiento, 1989) probably represent Cretaceous transfer faults. If a subduction-related magmatic arc existed at the current location of the Central Cordillera during the Cretaceous, as has been proposed (Aspden et al., 1987; Toussaint and Restrepo, 1989, 1994; Toussaint, 1995b), and the orientation of that magmatic arc and the extensional basins chelon has been preserved, their oblique orientation and en e pattern would suggest oblique slip extension with a right-lateral strike-slip component. However, the available data cannot rule out the hypothesis that some rift arms form acute angles to the dominant NNE-SSW trend in a pattern similar to aborted aulacogen rifts. If the inverse or thrust faults that now dene the eastern and western borders of the EC originally were Cretaceous normal faults with a strike-slip component, inverted during the Cenozoic, their geometry in map view would provide some information about Mesozoic extensional faults. Lateral changes of Mesozoic thickness suggest this is the case, at least for the master faults, which probably dened the regional extensional basin geometry. Adopting this hypoth ramo, La Salina, Bitu ima, Magesis, we posit that the Guaica faults represent original extensional or dalena, and Boyaca transtensional faults. Analog model experiments of oblique extension produce a similar map view fault pattern (e.g., Tron and Brun, 1991). NW-SE transfer faults and possible normal faults with this orientation, as interpreted by Ecopetrol (1994) in the middle MV, were not inverted during the Cenozoic. Some normal faults were passively transported with short-cut basement blocks during the Cenozoic inversion (e.g., Esmeraldas fault; Cooper et al., 1995). 9. Conclusions High-resolution backstripping analysis and forward modeling show that the Mesozoic Colombian Basin is marked by ve lithosphere stretching pulses. Periods of extension seem to correlate in time with gaps of subduction-related magmatic arc activity, as suggested by Aspden et al. (1987), especially during the Jurassic, which supports a hypothesis of backarc extension. However, this preliminarily correlation must be tested with more and better geochronological and biostratigraphical data. If backarc extension continued during the Early Cretaceous through oblique plate convergence, it probably has a strong strike-slip component. Evidence of a backarc basin located behind a partially emerged, less subsiding paleo-Central Cordillera (magmatic arc?) during the Late JurassicEarly Cretaceous includes the following: (1) Cretaceous plutonic bodies in the Central Cordillera; (2) Lower Cretaceous sandstones in the western part of the Cundinamarca subbasin with abundant volcanic lithic fragments and feldspar derived from a western detrital source area, indicated by tica sandpaleocurrent data (Murca Formation and U stone); (3) progressive westerly onlap terminations of the Cretaceous carbonates on the basement, observed in seismic lines, in the western border of the Cesar Valley in northern Colombia; (4) petrographical evidence suggesting

408

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

that Berriasian (?)Valanginian clastic sediments near San Felix in the western ank of the Central Cordillera came from the erosion of nearly uplifted areas containing fragments of metamorphic rocks and small tectonic blocks with plutonic rocks; (5) Cretaceous volaniclastic rocks that probably also derived from a magmatic arc; and (6) Late Cretaceous zircon ssion-track ages in the Central Cordillera. However, a passive margin hypothesis (Pindell and Erikson, 1993) or aborted rift arms related to the Caribbean opening cannot be completely ruled out because of the absence of a well-dened Cretaceous magmatic arc. Preliminarily, three stretching events are suggested during TriassicJurassic times. These events must be considered preliminary because the poor continental Triassic and Jurassic fossiliferous sedimentary record oers relatively scarce, low-quality biostratigraphic data that make dening clear time boundaries between events dicult. The spatial distribution of the lithosphere stretching values suggests that small, narrow (<150 km), asymmetric graben basins were located on opposite sides of the paleo-MagdalenaLa Salina fault system, which probably was active as a master transtensional or strike-slip fault system. Paleomagnetic data suggesting a signicant (at least 10) northward translation of terranes west of the Bucaramanga fault during the Early Jurassic (Bayona et al., 2005) and the similarity of early Mesozoic stratigraphy and the tectonic set terrane with the Late Permian ting of the Payande transtensional rift of the EC of Peru and Bolivia (Sempere et al., 2002) suggest that the areas may have been adjacent in early Mesozoic times. Additional geochronological, petrological, stratigraphic, and structural research should verify this hypothesis. In particular, additional paleomagnetic investigations should determine the paleolatitudinal position of the Central Cordillera and adjacent tectonic terranes during the TriassicJurassic. Subsidence curves, isopach maps, and paleomagnetic data (Bayona et al., 2005) suggest that during the BerriasianHauterivian, a (>180 km) wide, asymmetrical, transtensional half-rift basin existed, divided by the Santander Floresta high. A single depocenter in the south was limited at its southern reach by a vertical transfer fault. Small mac intrusions coincide with areas of thin crust (crustal stretching factors >1.4) and places of maximum stretching of the subcrustal lithosphere. During the Aptianearly Albian, the basin extended south in the UMV. Dierent crustal and subcrustal stretching values suggest either some lowermost crustal decoupling between crustal and subcrustal lithosphere or increased thermal thinning that aected the mantle lithosphere. Late Cretaceous subsidence was mainly driven by lithospheric cooling, water loading, and horizontal compressional stresses generated by collision of oceanic terranes in western Colombia. Triassic transtensional basins were narrow and increased in width during Triassic and Jurassic times. Cretaceous transtensional basins were wider than TriassicJurassic basins. During the Mesozoic, the strike-slip component gradually decreased at the expense of the

increase of the extensional component, according to paleomagnetic data (Bayona et al., 2005) and lithosphere stretching values. During the BerriasianHauterivian, the eastern side of the extensional basin may have developed by reactivation of an older Paleozoic rift system associat ramo fault system (Hossack et al., ed with the Guaica 1999). The western side probably developed through reactivation of an earlier normal fault system developed during TriassicJurassic transtension. Alternatively, the eastern and western margins of the graben may have developed along older strike-slip faults, the boundaries ramo fault of the accretion of terranes west of the Guaica during the Late Triassic and Jurassic (Bayona et al., 2005). During the rst stages of Cretaceous lithosphere extension, lithosphere necking started at a deep level, then evolved to shallower levels in latter extensional stages. The increasing width of the graben system likely was the result of progressive tensional reactivation of preexisting upper crustal weakness zones. Lateral changes of Mesozoic sediment thickness suggest that the reverse or thrust faults that now dene the eastern and western borders of the EC originally were normal faults with a strike-slip component that inverted during the Cenozoic ramo, La Salina, Andean orogeny. Thus, the Guaica were originally trans ima, Magdalena, and Boyaca Bitu tensional faults. The oblique orientation of most relative to the Mesozoic magmatic arc of the Central Cordillera may result from oblique slip extension during the Cretaceous or was inherited from the pre-Mesozoic structural grain. However, not all Mesozoic transtensional faults were inverted; some normal faults were passively transported with short-cut basement blocks during Cenozoic inversion. Thermal processes were more dominant than mechanical stretching during Late TriassicEarly Jurassic phases than the Cretaceous extensional phase. During the Late TriassicEarly Jurassic, abundant volcaniclastic rocks suggest a positive thermal anomaly in the lithosphere but moderate lithosphere stretching. TriassicJurassic age unconformities could have been produced by thermal uplift (active rifting?). In contrast, during the Cretaceous, less abundant volcanic rocks, the absence of tectonically controlled unconformities, and signicant tectonic subsidence indicates the absence of thermal doming. Minor mac intrusions coincident with places of maximum crustal and mantle subcrustal stretching suggest modest magmatism took place as a result of the extension of the lithosphere (passive rifting). Acknowledgments This research was part of the doctoral thesis of L.F. Sarmiento, supported in Colombia by the Icetex-Ecopetrol Fund, the Petroleum Colombian Company ECOPETROL, leo ICP and in the and the Instituto Colombiano del Petro Netherlands by the Netherlands Research School of Sedimentary Geology and the Tectonics Group of the Vrije

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411

409

Universiteit at Amsterdam. L.F. Sarmiento publishes with the permission of Ecopetrol-ICP. We acknowledge German Bayona, who shared his paleomagnetic results that complemented a previous version of this article and suggested improvements to the manuscript. This article also beneted from reviews by Franc ois Roure, Reini Zoetemeijer, Harry Doust, and Pedro Restrepo-Pace. References
a de la Cordillera Central y el Occidente Alvarez, J., 1983. Geolog mica de los intrusivos granitoides MesoceColombiano, y petroqu icos. Ph.D. Thesis Univ. Chile. Boletin Geolo gico Ingeominas, nozo 26(2), 1175. Bogota Aspden, J.A., McCourt, Brook, M., 1987. Geometrical control of subduction-related magmatism: the Mesozoic and Cenozoic plutonic history of Western Colombia. J. Geol. Soc. London 144, 893905. Audemard, F., 1991. Tectonics of Western Venezuela. Ph.D. Thesis. Rice University, Houston, TX, 243p. a, D.F., Mora, G., 1994. La Formacio n Saldan Bayona, G.A., Garc a: producto de la actividad de estratovolcanes continentales en un gicos dominio de retroarco. In: Etayo Serna, F. (Ed.), Estudios geolo del Valle Superior del Magdalena. Univ. Nacional de Colombia, , 21p., Chapter I. Ecopetrol, Bogota Bayona, G., Rapalini, A.E., Constanzo-Alvarez, V., Montes, C., Veloza, mez-Casallas, M., Silva, C., 2005. Mesozoic G., Ayala-Calvo, R.C., Go terrane translations and crustal block rotations in the Eastern Cordillera and Magdalena Valley of Colombia, as inferred from paleomagnetism. In: 6th International Symposium of Andean Geodynamics, 4p., submitted. Bessis, F., 1986. Some remarks on the study of subsidence of sedimentary basins; application to the Gulf of Lions margin (western Mediterranean). Marine Petroleum Geol. 3 (1), 3763. Bond, G., Kominz, M.A., 1984. Construction of tectonic subsidence curves for the early Paleozoic miogeocline, southern Canadian Rocky Mountains: implications for subsidence mechanisms, age of break-up, and crustal thinning. Geol. Soc. Am. Bull. 95, 155173. Bott, M.H.P., 1992. Passive margins and their subsidence. J. Geol. Soc. London 149, 805812. talloge nique des gisements Branquet, Y., 1999. Etude structurale et me ` l 0 histoire tectono-se meraude de Colombie: Contribution a dimend0e ` re Orientale de Colombie. The ` se de Doctorat de taire de la Cordille Geosciences, lInstitut National Polytechnique de Lorraine, Specialite ` res premie ` res et environement, Institut National Polytechnique Matie trographiques et ge ochimiques de Lorraine, Centre de recherches pe (CRPG/CNRS), 265p. gica de Colombia. Rev. Acad. Col. Cienc. Bu rgl, H., 1961. Historia geolo 9 (43), 137191. Exac. Fis. Nat. Bogota Bu rgl, H., 1967. The orogenesis of the Andean system of Colombia. Tectonophysics 4 (4-6), 429443. n, una molasa Mesozo ica de la Cordillera Cediel, F., 1968. El Grupo Giro 16 (13), 596. Oriental. Serv. Geol. Nal. Bol. Geol. Bogota ceres, C., 2003. Tectonic assembly of the Cediel, F., Shaw, R.P., Ca Northern Andean Block. In: Bartolini, C., Buer, R.T., Blickwede, J. (Eds.), The Circum-gulf of Mexico and the Caribbean: Hydrocarbon Habitats, Basin Formation, and Plate Tectonics. AAPG memoir 79, pp. 815848. Cloetingh, S., 1988. Intra-plate stress: a new element in basin analysis. In: Kleinspehn, K.L., Paola, C. (Eds.), Frontiers in sedimentary geology New perspectives in basin analysis. Springer-Verlag, New York, NY, pp. 205230. Cloetingh, S., Kooi, H., 1992. Intraplate stress and dynamical aspects of rift basins. In: Ziegler, P.A. (Ed.), Geodynamics of Rifting, Vol. II. Thematic Discussions. Tectonophysics 215, 167185. Cooper, M.A., Addison, F.T., Alvarez, R., Coral, M., Graham, R.H., Hayward, A.B., Howe, S., Martinez, J., Naar, J., Pen as, R., Pulham,

A.J., Taborda, A., 1995. Basin development and tectonic history of the Llanos Basin, Eastern Cordillera, and Middle Magdalena Valley, Colombia. A.A.P.G. Bull. 79 (10), 14211443. Duncan, R.A., Hargraves, R.B., 1984. Plate tectonic evolution of the Caribbean region in the mantle reference frame. In: Bonini, W.E., Hargraves, R.B., Shagam, R. (Eds.), The Caribbean-South American Plate Boundary and Regional Tectonics. Geol. Soc. Am., Mem. 162, 8193. Ecopetrol, Esso and Exxon Exploration Company, 1994, Integrated technical evaluation Santander sector Colombia, 19911994. Report, text and gures. Final Report, Houston, 38p., 49 gs. Etayo-serna, F., Renzoni, G., Barrero, D., 1976. Contornos sucesivos del cico en Colombia. In: Primer Congreso Colombiano de mar Creta a, Mem., Bogota , pp. 217252. Geolog gicos del Valle Superior del Etayo-Serna, F. (Ed.), 1994. Estudios geolo , 200p. Magdalena. Univ. Nacional de Colombia, Ecopetrol, Bogota tacico, Etayo-Serna, F., Laverde-Montan o, F. (Eds.), 1985. Proyecto Cre , 200p. contribuciones. Ingeominas, Publ. Geol. Esp. 16, Bogota lez, H., Etayo-Serna, F., Barrero, D., Lozano, H., Espinosa, A., Gonza lvarez, n Orrego, A., Zambrano, F., Duque, H., Vargas, R., Nu ez, A., A n, C., Ballesteros, I., Cardozo, E., Forero, H., Galvis, N., J., Ropa rez, C., Sarmiento, L., 1983. Mapa de terrenos geolo gicos de Ram , 235p. Colombia. Ingeominas Publ. Geol. Esp. 14, Bogota Fabre, A., 1983a. La subsidencia de la Cuenca del Cocuy (Cordillera ceo y el Terciario Inferior. Oriental de Colombia) durante el Creta a Primera parte: Estudio cuantitativo de la subsidencia. Geolog 8, 4961. Norandina, Bogota Fabre, A., 1983b. La subsidencia de la Cuenca del Cocuy (Cordillera ceo y el Terciario Inferior. Oriental de Colombia) durante el Creta n tecto nica. Geolog a Norandina, Segunda parte: Esquema de evolucio 8, 2127. Bogota mica de la sedimentacio n Creta cica en la regio n de la Fabre, A., 1985. Dina Sierra Nevada del Cocuy (Cordillera Oriental de Colombia). In: tacico, Etayo-Serna, F., Laverde-Montan o, F. (Eds.), Proyecto Cre , 20p. contribuciones. Chapter XIX, Ingeominas Publ. Esp. 16, Bogota ` le de neration dhydrocarbures: Un mode Fabre, A., 1987. Tectonique et ge ` re Orientale de Colombie et du Bassin des volution de la Cordille le ` ve 40 (Fasc. tace et le Tertiaire. Arch. Sci. Gene Llanos pendant le Cre 2), 145190. sicas Creta cicas de la Fabre, A., Delaloye, M., 1983. Intrusiones ba a. Norandina, Bogota 6, 1928. Cordillera Oriental. Geolog a de secuencias de Fajardo, A., Rubiano, J.L., Reyes, A., 1993. Estratigraf ceo Tard o al Eoceno Tard o en el sector central de las rocas del Creta la cuenca de los Llanos Orientales. Departamento del Casanare, Report ECP-ICP-001-93, Piedecuesta, Santander, 69p. z, Kennedy, Fo llmi, K.B., Garrison, R.E., Ramirez, P.C., Zabrano-Ort W.J., Lehner, B.L., 1992. Cyclic phosphate-rich successions in the Upper Cretaceous of Colombia. Palaeogeography, Palaeoclimatology, Palaeoecology 93, 151182. Geotec, 1992. Facies distribution and tectonic setting through the Phanerozoic of Colombia. A regional synthesis combining outcrop and subsurface data presented in 17 consecutive rock-time slices. , 100p. Bogota cas y faciales en el Tria sico Geyer, O., 1982. Comparaciones estratigra a. Norandina, Bogota 5, 2731. Norandino. Geolog mez, E., Pedraza, P., 1994. El Maastrichtiano de la regio n Honda Go mite N del Valle Superior del Magdalena: Registro Guaduas, l os trenzados. In: Etayo sedimentario de un delta dominado por r gicos del Valle Superior del Magdalena. Serna, F. (Ed.), Estudios geolo , 20p. Chapter III, Univ. Nacional de Colombia, Ecopetrol, Bogota mez, E., Jordan, T., Hegarty, K., Kelley, S., 1999. Diachronous Go deformation of the Central and Eastern Andean cordilleras of Colombia and syntectonic sedimentation in the Middle Magdalena Valley Basin. In: Fourth ISAG, Gottingen, Germany, 46 October, pp. 287290. Gradstein, F.M., Ogg, J.G., 1996. A Phanerozoic time scale. Episodes 9 (1-2), 35. Guerrero, J., Sarmiento, G., Navarrete, R.E., 2000. The stratigraphy of the W side of the Cretaceous Colombian Basin in the Upper

410

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411 Moreno, J.M., Concha, A.E., 1993. Nuevas manifestaciones gneas sicas en el anco occidental de la Cordillera Oriental, Colombia. ba a Colombiana 18, 143150. Geolog Moreno, M., Pardo, A., 2003, Stratigraphical and sedimentological constraints on Western Colombia: implications on the evolution of the Caribbean Plate. In: Bartolini, C., Buer, R.T., Blickwede, J. (Eds.), The Circum-gulf of Mexico and the Caribbean: Hydrocarbon Habitats, Basin Formation, and Plate Tectonics. AAPG memoir 79, 891924. ime and volcanic Nivia, A., 1987, The geochemistry and origin of the Ama sequences, SW Colombia. M. Phil. Thesis, Univ. of Leicester, UK, 164p. mez, J., 2005, Consideraciones acerca del modelo geolo gico Nivia, A., Go menes X evolutivo del Occidente Colombiano (Colombia). Resu a. Bogota , 2p. Congreso Colombiano de Geolog Nivia, A., Mariner, G., Kerr, A.C., 1996. El Complejo Quebradagrande nica del Creta ceo Inferior en la una possible cuenca marginal intracrato Cordillera Central de los Andes Colombianos. In: Paper presented to a, unpublished. the VII Congreso Colombiano de Geolog Pessagno, E.A., Martin, C., 2003. Tectonostratigraphic evidence for the origin of the Gulf of Mexico. In: Bartolini, C., Buer, R.T., Blickwede, J. (Eds.), The Circum-gulf of Mexico and the Caribbean: Hydrocarbon Habitats, Basin Formation, and Plate Tectonics. AAPG memoir 79, 4674. Pimpirev, C.T., Patarroyo, P., Sarmiento, G., 1992. Stratigraphy and facies analysis of the Caqueza Group, a sequence of Lower Cretaceous turbidites in the Cordillera Oriental of the Colombian Andes. Geol. 17, 297308. Colombiana, Bogota Pindell, J., Dewey, J., 1982. Permo-Triassic reconstruction of western Pangaea and the evolution of the Gulf of Mexico-Caribbean region. Tectonics 1, 179211. Pindell, J., Erikson, J., 1993. The Mesozoic margin of northern South America. In: Salty, J. (Ed.), Cretaceous tectonics of the Andes. Vieweg Germany, pp. 160. Pindell, J.L., Tabbutt, K.D., 1995. Mesozoic-Cenozoic Andean paleogeography and regional controls on hydrocarbon systems. In: Tankard, A.J., Suarez, R., Welsink, H.J. (Eds.), Petreoleum Basins of South America. A.A.P.G. Mem. 62, 101128. a, J.H., Rodr guez, O.G., 1978. Posibles turbiditas del Creta ceo Polan ) en el a rea de Anapoima (Cundinamarca); Inferior (Miembro Socota n sedimentolo gica basada en registros gra cos. Geol. una investigacio 10, 8791. Colombiana, Bogota Restrepo, J.J., Toussaint, J.F., Gonzalez, H., Gordani, U., Kawashita, K., gicas sobre el Linares, E., Parica, C., 1991. Precisiones geocronolo occidente Colombiano. Simposio sobre el magmatismo Andino y su nico. Programa Internacional de Correlacio n Geolo gica marco tecto gicas. Manizales, UNESCO, Union Internacional de Ciencias Geolo Colombia, Julio 35 de 1991. Manizales, pp. 121. Restrepo-Pace, P.A., 1995. Late Precambrian to Early Mesozoic tectonic evolution of the Colombian Andes, based on new geochronological, geochemical and isotopic data. Ph.D. Thesis. Univ. Arizona, 195p. Rivadeneira, M.V., 1996. Cretaceous-Paleogene stratigraphic sequences and the early Andean orogenic events in the Ecuadorian Oriente Basin. In: Third ISAG. St-Malo, France, 1719, September/1996, Extended abstracts, pp. 469472. guez, C., Rojas, R., 1985. Estratigraf a y tecto nica de la serie Rodr cica en los alrededores de San Fe lix, Cordillera Central de InfraCreta Colombia. In: Etayo-Serna, F., Laverde-Montan o, F. (Eds.), Proyecto cico, contribuciones. Ingeominas Publ. Esp. 16, Bogota , 21p., Creta Chapter XXI. Roeder, D., Chamberlain, R.L., 1995. Eastern Cordillera of Colombia: rez R., S., Jurassic-Neogene crustal evolution. In: Tankard, A.J., Sua Welsnik, H.J. (Eds.), Petroleum Basins of South America. A.A.P.G. Mem. 62, 633645. n, L.F., Carrero, M.M., 1995. Ana lisis estratigra co de la seccio n Rolo cica aorante al oriente del anticlinal de Los Cobardes entre los Creta n. Departamento de Municipios de Guadalupe-Chima-Contratacio a, Univ. Nacional de Colombia, Santander. Tesis pregrado Geolog , 80p. Bogota

Magdalena Valey. Revaluation of the select areas andtype localities a Colombiana. Bogota including Aipe, Guaduas and Piedras. Geolog 25, 45110. so-Ce nozo e interGuillande, M.R., 1988. Evolution Me que dune valle raine andine: La Haute Valle e du Rio Magdalena (Colombie). cordille Ph.D. Thesis Univer. Paris 6, Paris, 352p. ` re Orientale de Colombie brard, F., 1985. Les foothills de la Cordille He odynamique depuis entre les rios Casanare et Cusiana. Evolution ge tace . Ph.D. Thesis. Univ. Pierre et Marie Curie, Paris, 162p. dEo Cre Hill, R.I., 1993. Mantle plumes and continental tectonics. Lithos 30, 193 206. nez, J., Estrada, C., Herbert, R., 1999. Structural Hossack, J., Mart evolution of the Llanos fold and thrust belt, Colombia. In: McClay, K. (Ed.), Thrust Tectonics 99 Meeting. Royal Halloway Univ. of London, April 2629, June 1999, Program, London, 110p. Ingersoll, R.V., Busby, C.J., 1995. Tectonics of sedimentary basins. In: Busby, C.J., Ingersoll, R.V. (Eds.), Tectonics of Sedimentary Basins. Blackwell Science, pp. 15. n de las edades de algunas rocas de la Jaramillo, J.M., 1978. Determinacio todo de huellas de sio n. Cordillera Central de Colombia por el me , pp. 1920. menes, Bogota 2do. In: Congreso Colomb. Geol. Resu n de las edades de algunas rocas de la Jaramillo, J.M., 1981. Determinacio todo de huellas de sio n. Bol. Cordillera Central de Colombia por el me n 56, 145146. Cienc. de la Tierra, Medell Julivert, M., 1968. In: Colombie. Lexique stratigraphique International. Precambrian, Paleozoique et Mesozoique. Paris, vol. V, Fasc. 4a, premiere partie, 651p. Kammer, A., 1993. Steeply dipping basement faults and associated structures of the Santander Massif, Eastern Cordillera, Colombian 18, 4764. Andes. Geol. Colombiana, Bogota mez, L., Tilander, N., 1997. Kluth, Ch., F., Laad, R., De Armas, M., Go Dierent structural styles and histories of the Colombian foreland: Castilla and Chichimene oil elds areas, east-central Colombia. In: VI Simpolsio Bolivariano Exploracion en las Cuencas Subandinas, Cartagena de Indias, Mem. Tomo II, pp. 185197. Kooi, H., Burrus, J., Cloetingh, S., 1992. Lithospheric necking and regional isostasy at extensional basins: part 1, Subsidence and gravity modeling with an application to the Gulf of Lions margin (SE France). J.G.R. 97, 1755317571. Linares, R., 1996. Structural styles and kinematics of the Medina area, Eastern Cordillera, Colombia. M.Sc. Thesis, Univ. of Colorado, 104p. Litherland, M., Aspden, J.A., Jamielita, R.A., 1994. The metamorphic belts of Ecuador. British Geological Survey, Overseas Memoir. 11, 1146. , Maze, W.B., 1984. Jurassic la Quinta Formation in the Sierra de Perija northwestern Venezuela: geology and tectonic environment of red beds and volcanic rocks. In: Bonini, W.E., Hargraves, R.B., Shagam, R. (Eds.), The Caribbean-South American Plate Boundary and Regional Tectonics. Geol. Soc. Amer. Mem. 162, 263282. McCourt, W.J., Feininger, T., Brook, M., 1984. New geological and geochronological data from the Colombian Andes: continental growth by multiple accretion. J. Geol. Soc. London 141, 831845. McKenzie, D., 1978. Some remarks on the development of sedimentary basins. Earth Planet. Sci. Lett. 40, 2532. Area, McLaughlin Jr., D.H., 1972. Evaporite deposits of Bogota Cordillera Oriental, Colombia. A.A.P.G. Bull. 56 (11), 22402259. Meschede, M., Frisch, W., 1998. A plate tectonic model the Mesozoic and Early Cenozoic history of the Caribbean Plate. Tectonophysics 296 (3 4), 269291. sico del sector norocciMojica, J., Kammer, A., Ujueta, G., 1996. El Jura a de la excursio n al Valle Superior del dental de Suramerica y gu y Prado. DepartamenMagdalena (Nov. 14/95), Regiones de Payande , 21p. to del Tolima, Colombia. Geol. Colombiana, Bogota Moreno, J.M., 1990. Stratigraphy of the Lower Cretaceous units central part Eastern Cordillera, Colombia. In: 13th International Sedimentological Congress, Nothingham, August 2231, 1990. Abstract. Moreno, J.M., 1991. Provenance of the Lower Cretaceous sedimentary sequences, central part, Eastern Cordillera, Colombia. Rev. Acad. Col. Cienc. Exact. Fis. y Nat. 18 (69), 159173.

L.F. Sarmiento-Rojas et al. / Journal of South American Earth Sciences 21 (2006) 383411 Ross, M.I., Scotese, C.R., 1988. A hierarchical tectonic model of the Gulf of Mexico and Caribbean region. Tectonophysics 155, 139168. Royden, L.H., 1993a. The tectonic expression of slab pull at continental convergent boundaries. Tectonics 12, 303325. Royden, L.H., 1993b. Evolution of retreating subduction boundaries formed during continental collision. Tectonics 12, 629683. Royden, L., Keen, C.E., 1980. Rifting process and thermal evolution of the continental margin of eastern Canada determined from subsidence curves. Earth Planet. Sci. Lett. 51, 343361. Rubiano, J.L., 1989. Petrography and stratigraphy of the Villeta Group, Codillera Oriental, Colombia, South America. M.Sc. Thesis, Univ. South Carolina, Columbia, SC, 96p. Sarmiento, L.F., 1989. Stratigraphy of the Cordillera Oriental west of , Colombia. M.Sc. Thesis University of South Carolina, Bogota Columbia, SC, 102p. Sarmiento, L.F., 2001. Mesozoic rifting and Cenozoic basin inversion history of the Eastern Cordillera, Colombian Andes. Inferences from tectonic models. Ph.D. Thesis Vrije Universiteit, Amsterdam, The Netherlands, 297p. Sclater, J.G., Christie, P.A.F., 1980. Continental stretching; an explanation of the post mid-Cretaceous subsidence of the central North Sea Basin. J.G.R. 85, 37113739. Sempere, T., Carlier, G., Soler, P., Fornari, M., Carlotto, V., Jacay, J., raudeau, D., Cardenas, J., Rosas, S., Jime nez, N., 2002. Arispe, O., Ne Late Permian-Middle Jurassic lithospheric thinning in Peru and Bolivia, and its bearing on Andean-age tectonics. Tectonophysics 345, 153181. Shagam, R., 1975. The Northern termination of the Andes. In: Nairn, A.E.M., Stehli, H. (Eds.), The Ocean Basins and Margins, vol. 3. The Gulf of Mexico and the Caribbean. Plenum Press, New York and London, pp. 325420, Chapter 9. Smith, G.A., Landis, C.A., 1995. Intra-arc basins. In: Busby, C.J., Ingersoll, R.V. (Eds.), Tectonics of sedimentary basins. Blackwell Science, pp. 263298. Steckler, M.S., Watts, A.B., 1978. Subsidence of the Atlantic-type continental margin o New York. Earth Planet. Sci. Lett. 41, 113. Toro, G., Popeau, G., Hermelin, M., Schwabe, E., 1999. Chronology of the volcanic activity and regional thermal events: a contribution from the tephrochronology in the north of the Central Cordillera, Colombia. In: Fourth ISAG. Geottingen, Germany, 46 October/1999, Extended Abstracts, pp. 761763. tesis sobre el marco geodina mico de Colombia Toussaint, J.F., 1995a. Hipo ico temprano, Contribution to IGCP 322 Jurassic durante el Mesozo 20, 150155. events in South America. Geol. Colombiana, Bogota n geolo gica de Colombia 2. Tria sico Toussaint, J.F., 1995b. Evolucio sico. Contribucio n al IGCP 322 Correlation of Jurassic events in Jura South America International Geological Correlation Programme n, 94p. Unesco IUGS. Univ. Nacional de Colombia. Medell Toussaint, J.F., Restrepo, J.J., 1989. Acresiones sucesivas en Colombia; n geolo gica. V Congr. Colomb. Geol., un nuevo modelo de evolucio Bucaramanga I, 127146.

411

Toussaint, J.F., Restrepo, J.J., 1994. The Colombian Andes during Cretaceous times. In: Salty, J.A. (Ed.), Cretaceous tectonics of the Andes. Views Brunwick, Germany, pp. 61100. Tron, V., Brun, J.P., 1991. Experiments on oblique rifting in brittle-ductile systems. Tectonophysics 188, 7184. guez, E., 1976. Geolog a del cuadra ngulo K-12. GuateUlloa, C., Rodr 22 (1), 455. que. Bol. Geol., Ingeominas, Bogota Van der Beek, P., 1995. Tectonic evolution of continental rifts, inferences from numerical modeling and ssion track thermochronology. Ph.D. Thesis, Vrije Universitteit, Amsterdam, 232p. Van der Hamen, T.H., 1961. Late Cretaceous and Tertiary stratigraphy and tectogenesis of the Colombian Andes. Geologie en Mijnbouw 40, 181188. Van der Wiel, A.M., 1991. Uplift and volcanism of the SE Colmbian Andes in relation to Neogene sedimentation of the Upper Magdlena Valley. Ph.D. Thesis, Wageningen, 208p. Van Wees, J.D., Stephenson, R.A., Stovba, S.M., Shymanovskyi, V.A., 1996b. Tectonic variation in the Dnieper-Donets Basin from automated modeling of backstripped subsidence curves. Tectonophysics 268 (1-4), 257280. Formation (Aptian Vergara, L., Pro ssl, K.F., 1994. Dating the Yav Upper Magdalena Valley, Colombia). In: Etayo Serna, F. (Ed.), gicos del Valle Superior del Magdalena. Univ. Nacional Estudios geolo , 14p., Chapter XVIII. de Colombia, Ecopetrol, Bogota Vesga, C.J., Barrero, D., 1978. Edades K/Ar en rocas gneas y rcas de la Cordillera Central de Colombia y su implicacio n metamo gica. II Congreso Colombiano de Geolog a, Resu menes, geolo Bogota. Villamil, T., 1993. Relative sea level, chronology, and a new sequence stratigraphy model for distal oshore facies, Albian to Santonian, Colombia. In: Pindell, J.A., Drake, C.D. (Eds.), Mesozoic-Cenozoic Stratigraphy and Tectonic Evolution of the Caribbean Region/ Northern South America: Implications for Eustasy from Exposed Sections of a Cretaceous-Eocene Passive Margin Setting. Geol. Soc. Amer. Memoir, paper C-8. Villamil, T., 1994. High-resolultion stratigraphy, chronology and relaitve sea level of the Albiian-Santonian (Cretaceous) of Colombia. Ph.D. Thesis Univ. Of Colorado at Boulder, 446p. Villamil, T., Arango, C., 1998. Integrated stratigraphy of latest Cenomanian and early Turonian facies of Colombia. Paleogeographic evolution and non-glacial eustasy, Northern South America. SEPM Spec. Publ. 58, 129159. Watts, A.B., Karner, G.D., Steckler, M.S., 1982. Lithospheric exure and the evolution of sedimentary basins. Phil. Trans. Roy. Soc. London 305, 249281. Wilson, M., 1993. Magmatism and the geodynamics of basin formation. Sediment. Geol. 86, 529. Ziegler, P.A., 1994. Geodynamic processes governing development of rifted basins. In: Roure, F., Ellouz, N., Shein, V.S., Skvortsov, I. (Eds.), Geodynamic Evolution of Sedimentary Basins. Institut Franc trole, Technip, Paris, pp. 1697. ais du Pe