Episodic orogenic front migration in the northern Andes:

Constraints from low-temperature thermochronology in the

Eastern Cordillera, Colombia

Mauricio Parra,1 Andres Mora,1,2 Edward R. Sobel,1 Manfred R. Strecker,1

and Roman Gonzalez3

Received 6 November 2008; revised 17 March 2009; accepted 9 April 2009; published 10 July 2009.

[1] New thermochronometric data from the EasternCordillera of the Colombian Andes reveal diachronousexhumation associated with Cenozoic contractionaldeformation in this sector of the northern Andes. Wepresent a comprehensive account of exhumationpatterns along a 150-km-long, across-strike transectbetween �4� and 6�N by integrating 29 new apatitefission track (AFT) ages and 17 new zircon fission track(ZFT) ages with sparse published thermochronologicaldata from this area. Our data reveal episodic eastwardmigration of the orogenic front at an average rate of2.5–2.7 mm/a during the Late Cretaceous-Cenozoic.We identify three major stages of orogen propagation:(1) slow propagation (0.5–3.1 mm/a) until earlyEocene; (2) rapid orogenic advance (4.0–18.0 mm/a)during middle-late Eocene, which accounts for �86%of the orogen’s present width; and (3) slow orogenpropagation (1.2–2.1 mm/a) from Oligocene toHolocene times. Our data demonstrate that in the courseof changes in plate kinematics, the presence of inheritedcrustal anisotropies, such as the former rift-boundingfaults of the Eastern Cordillera, favor a nonsystematicprogression of foreland basin deformation through timeby preferentially concentrating accommodation of slipand thrust loading along these zones of weakness.Citation: Parra, M., A. Mora, E. R. Sobel, M. R. Strecker, and

R. Gonzalez (2009), Episodic orogenic front migration in the northern

Andes: Constraints from low-temperature thermochronology in

the Eastern Cordillera, Colombia, Tectonics, 28, TC4004,

doi:10.1029/2008TC002423.

1. Introduction

[2] The large-scale kinematics of faulting in thrust beltshas been successfully reproduced by theoretical and analogmodels that consider that deformation of such belts resemblesthat of critically tapered wedges [e.g., Davis et al., 1983;

Dahlen et al., 1984; Dahlen and Barr, 1989; Dahlen, 1990;Hoth et al., 2005]. While successfully applied to isotropicmaterials [e.g.,Dahlen, 1990;DeCelles andMitra, 1995], thegeometry of deformation in thrust belts developed overprestrained segments of continental crust may be stronglyguided by preexisting anisotropies that absorb and guideupper crustal deformation [e.g., Allmendinger et al., 1983;Jordan and Allmendinger, 1986; Marshak et al., 2000;Pfiffner et al., 2000]. The flexural response of the lithosphereto the topographic load exerted by this kind of thick-skinnedthrust belt leads to the formation of broken foreland basins, inwhich the preexisting anisotropies determine the areal extentand compartmentalization of an otherwise contiguous basin[e.g., Jordan, 1981; Allmendinger et al., 1983; Jordan andAllmendinger, 1986; Streckler et al., 2009]. As convergenceand deformation progress, a nonsystematic, spatially dispa-rate and diachronous pattern of orogenic advance and incor-poration of these depocenters into the fold-and-thrust belt isexpected for these orogens [e.g.,Hilley et al., 2005; Streckleret al., 2009].[3] In orogens with moderate exhumation where ancient

sedimentary basins are still partially preserved in the orogeninterior, erosion may have resulted in a low degree ofpreservation of indicators of brittle deformation and rockuplift, such as growth strata or crosscutting structural rela-tions [e.g., Sobel and Strecker, 2003]. In the absence of suchindicators, however, quantification of the timing of rangeuplift relies either on ductile deformation markers [Hubbardand Harrison, 1989; Godin et al., 2001] or on indirectmethods, such as reconstructing the subsidence history ofadjacent basins [e.g., Echavarrıa et al., 2003; Jones et al.,2004]. In areas with moderate exhumation, direct quantifica-tion of long-term denudation may be limited to the youngestexhumation history of the uppermost crustal levels, whoseevolution can only be evaluated through low-temperaturethermochronology methods, such as apatite (U-Th)/He [e.g.,Crowhurst et al., 2002; Bertotti et al., 2006; Richardson etal., 2008] or apatite fission track dating, especially whencombined with other thermal history indicators such asvitrinite reflectance [Bray et al., 1992; O’Sullivan, 1999].In the case where thermochronometers were not reset duringburial, unraveling older exhumation events often requiresindirect, less precise methods, such as inverse modeling ofapatite fission track data [e.g., Steinmann et al., 1999].[4] Here, we present a case study of late Cretaceous-

Cenozoic episodic lateral growth of an inversion orogen inthe northern Andes of Central Colombia. In this setting,

TECTONICS, VOL. 28, TC4004, doi:10.1029/2008TC002423, 2009ClickHere

for

FullArticle

1Institut fur Geowissenschaften, Universitat Potsdam, Golm, Germany.2Instituto Colombiano del Petroleo, Ecopetrol, Bucaramanga, Colombia.3Hocol S.A., Bogota, Colombia.

Copyright 2009 by the American Geophysical Union.0278-7407/09/2008TC002423$12.00

TC4004 1 of 27

Cenozoic inversion and moderate exhumation of a Mesozoicrift basin that occupied the area of the present-day EasternCordillera between 4� and 5�300N (Figure 1) offers a uniqueopportunity to study the style of orogenic growth in asubduction orogen developed over a region of significantlyprestrained crust in a foreland setting. Here, the closeproximity of various morphotectonic provinces characterizedby different degrees of exhumation over an across-strikedistance of only �150 km, constitutes an ideal, yet poorlystudied, natural experimental setup, where rock uplift andexhumation can be evaluated. These morphotectonic prov-inces include (1) a slightly exhumed Cenozoic central foldbelt, including the 2.5-km-high intraorogenic Bogota pla-teau; (2) the Quetame and Floresta massifs constitute deeplyexhumed basement highs located to the east and northeast ofthe plateau, respectively, where Paleozoic, low- andmedium-grade metamorphic continental basement rocks, and Creta-ceous rift-related sedimentary rocks were uplifted throughcontractional reactivation of west dipping reverse faults[Kammer and Sanchez, 2006; Mora et al., 2006, 2009];(3) the moderately exhumed Villeta Anticlinorium, locatedto the west of the Bogota Plateau, where west verging reversefaults caused uplift and exhumation of Lower Cretaceous rift-related sedimentary units; and (4) fold-and-thrust belts oneither side, involving basins with up to 7 km of Cenozoicforeland deposits.[5] Available data (see below) on the spatiotemporal

tectonic evolution of the northern Andes indicate a LateCretaceous age for initiation of deformation in the ColombianAndes, in areas that constitute the present-day Central Cor-dillera [Gomez et al., 2003]. A Paleocene-early Oligoceneeastward advance of the orogenic front toward the EasternCordillera is supported by crosscutting relationships andgrowth strata in the western and axial sectors of the EasternCordillera [Gomez et al., 2003; Restrepo-Pace et al., 2004;Gomez et al., 2005], and by an increase in tectonic subsidencein basins to the east [Parra et al., 2009] (see section 3).However, despite this information, little is known aboutwhether deformation was coupled with exhumation, due tothe lack of thermochronologic studies constraining the earlydenudational history. AFT ages and thermal modeling fromthe western flank of the Eastern Cordillera illustrate an oldercooling event that began sometime between 65 and 30 Ma,and a younger cooling episode between 15 and 5Ma [Gomez,2001; Gomez et al., 2003] (see Table S1 in the auxiliarymaterial).1 On the other hand, published AFT data fromthe eastern flank of the Eastern Cordillera indicate that(1) exhumation had already begun by �20 Ma in theSantander Massif [Shagam et al., 1984] and (2) rapid exhu-mation of 3 to 5 km of crust has been active during the last 3–4 Ma [Mora et al., 2008]. Thus, these data do not furnishinformation on the earliest exhumation history of the range.[6] In this study, our primary goals are (1) to investigate

the spatial and temporal pattern of exhumation in theEastern Cordillera across a �150-km-long transect byintegrating 46 new AFT and ZFT analyses with publishedAFT and ZFT data; (2) to evaluate the temporal correlation

between these exhumation patterns with published directand indirect indicators of rock uplift throughout the Centraland Eastern cordilleras; and (3) to correlate the uplift andexhumation patterns with kinematic models for the evolu-tion of the Eastern Cordillera.

2. Regional Setting, Stratigraphy,

and Tectonic Evolution

[7] In the course of an absolute westward movement ofthe South American Plate over the Pacific ocean basin[Coney and Evenchick, 1994; Russo and Silver, 1996],contractional deformation along the tectonically active north-western margin of the South American continent results fromoblique convergence and accretion of oceanic crust of Pacificaffinity since late Cretaceous time [e.g., Cooper et al., 1995;Taboada et al., 2000]. Late Cretaceous initial accretion ofoceanic crust along the western margin of Ecuador andColombia [e.g., Spikings et al., 2001; Kerr and Tarney,2005; Vallejo et al., 2006; D. Villagomez et al., Thermotec-tonic history of the Northern Andes, paper presented at 7thInternational Symposium on Andean Geodynamics, Institutde Recherche pour le Developpement, Nice, 2008] triggeredinitial orogeny in the northern Andes [e.g., McCourt et al.,1984]. Regional kinematic reconstructions of the Farallon-Nazca and South American plates illustrate a major change inconvergence direction from a N-S to a �WNW-ESE at circa49 Ma [Pilger, 1984; Pardo-Casas and Molnar, 1987].Associated with this kinematic regime, paleostress determi-nations in the eastern Colombian Andes indicate a coevalshift in the direction of the main horizontal stress, whichrotated clockwise from a SW-NE orientation prevailing untilearly Eocene to aWNW-ESE direction during middle Eocene-Miocene time [Cortes et al., 2005], similar to the neotectonicorientation [Taboada et al., 2000; Colmenares and Zoback,2003; Dimate et al., 2003; Cortes and Angelier, 2005;Mora,2007]. The protracted plate convergence along the westernmargin of South America has generated a Cenozoic orogen inthe northern Andes which comprises three NE-SW strikingranges north of 2�N. These include (1) the Western Cordil-lera, formed by relicts of oceanic basement sliced off anoceanic plateau associated with a Late Cretaceous mantleplume [e.g., Burke, 1988; Kerr and Tarney, 2005]; (2) theCentral Cordillera, an exhumed, Mesozoic to Recent mag-matic arc that intruded Proterozoic and Paleozoic continentalbasement; and (3) the Eastern Cordillera, an inverted Meso-zoic rift with east and west dipping reverse faults, whichabuts the foredeep basin that extends eastward toward theGuyana craton [e.g., Cooper et al., 1995].[8] Structural models illustrate that at �4–5�N the main

range-bounding faults of the Eastern Cordillera merge into asubhorizontal detachment at a depth between 20 and 25 km[e.g., Restrepo-Pace et al., 2004; Cortes et al., 2006; Moraet al., 2006] (Figure 1b). Mora et al. [2006, 2008] inferreda midcrustal detachment in the Eastern Cordillera based onstructural area balancing and the presence of a zone offocused seismicity. These authors model the Servita faultwith a listric geometry because of the documented exhuma-tion pattern, which is reminiscent of that simulated for

1Auxiliary materials are available in the HTML. doi:10.1029/2008TC002423.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

2 of 27

TC4004

Figure 1

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

3 of 27

TC4004

inverted listric normal faults [e.g., Erslev, 1986]. Shorteningestimates in their model amount to 58 km. However, thisvalue should be considered as a conservative end-membercalculation for the Eastern Cordillera, in contrast to modelsarguing for higher shortening values (e.g., 150 km [Dengoand Covey, 1993] and 200 km [Roeder and Chamberlain,1995]), where the hanging wall transport above subhorizon-tal detachments occurs over longer distances. While the lattermechanism may produce more crustal thickening whenmultiple crustal-scale sheets are involved, it does not honormost of the observations supporting the listric fault geometryof Mora et al. [2008].[9] Along the axial and eastern sectors of the Eastern

Cordillera, in the Quetame and Floresta massifs, the exhumedbasement constitutes up to�4 km of low- and medium-grademetamorphic rocks and Paleozoic intermediate to acid intru-sives [Segovia, 1965; Ulloa and Rodrıguez, 1979; C. Ulloaand E. Rodrıguez, Intrusivos acidos ordivıcicos y post-devonicos en La Floresta (Boyaca), paper presented at IVCongreso Colombiano de Geologıa, Sociedad Colombianade Geologıa, Cali, 1982]. The disconformably overlyingrocks of late Paleozoic age comprise up to�4 km of a clasticplatformal sequence, including the Devonian Tibet, Floresta,and Cuche formations in the Floresta area [Mojica andVillarroel, 1984] and theDevonian to Carboniferous FarallonesGroup in the Quetame Massif [Ulloa and Rodrıguez, 1979].The inverted crust in the Eastern Cordillera has been heavilyinfluenced by Mesozoic rifting associated with two maintranstensional events during the Late Triassic-Middle Juras-sic and latest Jurassic-Cretaceous time [e.g., Kammer andSanchez, 2006; Sarmiento-Rojas et al., 2006]. Triassic-Middle Jurassic rift-related accumulation of mainly non-marine and volcanoclastic sediments occurred in narrowasymmetric graben straddling the western flank of the present-day Eastern Cordillera from the southern Magdalena Basin inthe south to the SantaMartaMassif in the north [e.g.,Sarmiento-Rojas et al., 2006, and references therein]. In the FlorestaMassif, at �6�N, a minimum of up to 1.8 km of Lower toUpper Jurassic lacustrine and fluvial sediments with tuffa-ceous interbeds accumulated within half graben basinslimited by west dipping faults [Kammer and Sanchez,2006]. These units comprise the Palermo, Montebel, andRusia formations in the western hanging wall of the Boyacafault, and the Giron Formation in the hanging wall of theSoapaga fault [Kammer and Sanchez, 2006] (Figure 2).Unlike the narrow geometry of the early Mesozoic depo-centers, regional reconstructions of the extent of LowerCretaceous sediments in the Eastern Cordillera reveal thatthe second Mesozoic rifting episode generated a widerextensional basin system bounded by the east dippingLa Salina fault system, to the west, and the west dipping

Servita-Lengupa fault system, to the east [e.g., Mora et al.,2006; Sarmiento-Rojas et al., 2006]. Along the footwall ofthese normal faults, rift shoulder erosion led to the partial ortotal removal of Paleozoic sequences [e.g., Mora et al.,2006]. The resulting paleotopography was preserved beneathsubsequent Lower Cretaceous shallow marine units, depos-ited during the active stage of rifting. As a result, drasticfacies and thickness changes (4.5–6 km) characterize theLower Cretaceous sediments [Mora et al., 2006, 2009] (seeFigure 2). These units are covered by �1.5–2 km of UpperCretaceous units (Une and Chipaque formations and theGuadalupe Group), deposited during a stage of postriftthermal subsidence [Sarmiento-Rojas et al., 2006].[10] Associated with deformation and crustal thickening

in the Central Cordillera since the latest Cretaceous, aforeland basin system began to develop east of this range,in areas that include the present-dayMagdalena Valley Basin,the Eastern Cordillera and the Llanos Basin [e.g., Cooper etal., 1995; Gomez et al., 2005]. Continued Cenozoic conver-gence and deformation prompted eastward orogenic growthand caused tectonic inversion of the Mesozoic rift basin [e.g.,Cooper et al., 1995; Mora et al., 2006; Bayona et al., 2008;Parra et al., 2009]. This initial deformation and uplift in theEastern Cordillera fragmented the foreland basin, shifted theforedeep eastward, to the region occupied by the modernwedge top basins of the eastern foothills, and ultimatelyresulted in the formation of intermontane basins in theBogota Plateau and the Magdalena Valley (Figure 1). TheCenozoic sedimentary record associated with foreland basindevelopment comprise a Paleocene to late Miocene coarsen-ing upward sequence, approximately 6.5 km thick (Figure 2).In the Medina Basin, along the eastern foothills of the range(see Figure 1 for location), this sequence evolves from coastalplain and tidally influenced lacustrine deposits of the Barco,Los Cuervos, Mirador, Carbonera, and Leon formations, tothe distal and proximal alluvial deposits of the GuayaboGroup [e.g., Cooper et al., 1995; Parra et al., 2009].Paleocurrent directions and sandstone petrography suggestthat the main sediment source for the lower Oligocene toMiocene deposits coincides with the present-day EasternCordillera [Bayona et al., 2008; Parra et al., 2009].

3. Deformation and Exhumation in the

Central and Eastern Cordilleras

[11] In this section we provide a brief synopsis ofavailable data on the chronology of deformation and exhu-mation in the central Colombian Andes. Provenance dataand sedimentary facies distribution along the western fold-and-thrust belt of the Eastern Cordillera and the Magdalena

Figure 1. (a) Geologic map of the Eastern Cordillera modified from Gomez et al. [2007],Mora et al. [2006], and Parra etal. [2009] showing main structures, morphotectonic domains, and location of fission track samples. Symbols are WC,Western Cordillera; CC, Central Cordillera; MMV, Middle Magdalena Valley Basin; VA, Villeta Anticlinorium; BP, BogotaPlateau; QM, Quetame Massif; FM, Floresta Massif; MB, Medina Basin; LIB, Llanos Basin. Boxes denote location offour fission track transects (Western (W), Northern (N) Central (C), and Southern (S) transects) shown in Figures 3–6.(b) Structural cross section showing the former rift-bounding La Salina and Servita-Lengupa faults, reactivated as reversefaults during Cenozoic deformation [after Mora et al., 2008].

4 of 27

TC4004TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

Figure 2. Generalized stratigraphy of the western, axial and eastern sectors of Eastern Cordillera. Thestratigraphic locations of fission track samples are indicated. Inset map denotes approximate line ofcorrelation.

5 of 27

TC4004TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

Valley Basin have been used to constrain the onset ofcontractional deformation of the Central Cordillera to lateCretaceous time (Campanian to Maastrichtian, �80–65 Ma[Cooper et al., 1995;Gomez et al., 2003; Ramon and Rosero,2006]). Associated with this episode of mountain building,ongoing exhumation since �80 Ma has been documentedwith a limited number of bedrock apatite and zircon fissiontrack data from the Central Cordillera [Gomez et al., 2003].To the east of this range, the timing of Cenozoic deformationalong a NW-SE transect at 4�–6�N has been determinedusing various indicators within a biostratigraphically datedforeland basin record [Gomez et al., 2003; Restrepo-Pace etal., 2004; Gomez et al., 2005; Bayona et al., 2008; Parra etal., 2009]. These data illustrate the Cenozoic eastwardadvance of the deformation front from its late Cretaceousposition in the Central Cordillera toward the present-dayEastern Cordillera. First, along its western fold-and-thrustbelt, regional biostratigraphic correlation and surface cross-cutting relations provided by an angular unconformity at thebase of middle Eocene nonmarine units [Gomez et al., 2003;Restrepo-Pace et al., 2004] constrain the age of initialdeformation to late Paleocene-middle Eocene (�56–43 Ma).Second, growth strata imaged on industry-style seismicreflection profiles along the eastern margin of the Guaduassyncline document initial nucleation of the Villeta Anticli-norium through contractional reactivation of the extensionalLa Salina fault system during the middle Eocene-earlyOligocene (43–30 Ma [Gomez et al., 2003]). Farther east,middle Eocene-lower Oligocene (48–30 Ma) growth stratasouth of the Bogota Plateau [Julivert, 1963; Gomez et al.,2005] reflect initial deformation in the axial Eastern Cor-dillera. Finally, even farther to the east, in the Medina Basin,detailed palynological data document a late Oligocene(�31 Ma) increase in subsidence rates, which has beenassociated with initial thrust loading in the Eastern Cordillera[Parra et al., 2009].[12] Despite this regional-scale information on the late

Cretaceous-Oligocene spatiotemporal pattern of deforma-tion, details about the Neogene pattern of orogenic growthas well as on the exhumation patterns associated with theeastward advance of the Colombian Andes are poorlyknown. Published thermochronological data include thermalmodeling of apatite fission track data from the Guaduassyncline, in the Eastern Cordillera’s western foothills, whichhas been used to suggest initiation of thrust-driven exhu-mation sometime between 65 and 30 Ma [Gomez et al.,2003]. In the Floresta Massif, late Oligocene-early Mioceneexhumation has been suggested on the basis of a limitednumber of AFT ages [Toro, 1990; Toro et al., 2004].Approximately �100–150 km to the north, in the SantanderMassif, AFT ages obtained in Proterozoic and Triassicintrusives are as old as 20 Ma, suggesting that exhumationin the axial sector and eastern margin of the EasternCordillera had already begun in the early Miocene [Shagamet al., 1984]. Finally, in the Quetame Massif, AFT agesyounger than 4 Ma reveal rapid Pliocene-Pleistocene ex-humation [Mora et al., 2008].[13] These available thermochronological data provide an

initial, yet incomplete picture of the exhumation of the

range. This is particularly the case in the axial and easternsectors, where the age for initiation of thrust-inducedexhumation is unknown.

4. Methods

[14] In order to unravel the early exhumation history ofthe Eastern Cordillera, we employed apatite and zirconfission track thermochronology (AFT and ZFT, respectively;see Tagami and O’Sullivan [2005] for a detailed descriptionof the method). The range of temperatures at which fissiontracks are progressively shortened defines the partial anneal-ing zone (PAZ). For apatite, it typically corresponds to �60�to 110–140�C [e.g., Ketcham et al., 1999]. The uppertemperature limit of the PAZ, referred as to the total annealingtemperature, depends on the cooling rate and has been shownto correlate with both the chemical composition of theapatites [e.g., Green et al., 1986] and the etching parameterDpar [Donelick et al., 1999;Ketcham et al., 1999;Donelick etal., 2005]. Commonly, apatites with larger Dpar are chlorinerich and anneal at higher temperatures [Donelick et al., 2005].In the case of ZFT, the fission track annealing kinetics largelydepends on the degree of radiation damage of zircon crystals[Rahn et al., 2004; Garver et al., 2005]. Accordingly,estimates of the total annealing temperature for ZFT are inthe range of 250 ± 40�C [e.g., Brandon et al., 1998; Bernetand Garver, 2005].[15] In the Eastern Cordillera of Colombia, spatially

uneven burial related to the Mesozoic rifting history causedlaterally variable peak burial temperatures. Our samplingstrategy was guided by paleoburial temperature estimatesderived from published vitrinite reflectance (R0) analyses(Table S2). In the axial part of the Eastern Cordillera at thelatitude of our transect, (R0) values of 0.6–0.7% occur inthe Upper Cretaceous sedimentary units (GuadalupeGroup). For normal heating rates of 1�–5�C/Ma, these R0

values are equivalent to the total annealing temperatures ofthe AFT system according to the kinetic model of Burnhamand Sweeney [1989]. Within the Lower Cretaceous synriftsediments of the Macanal Formation, �5–6 km deeperwithin the stratigraphic section, R0 values >3.5% indicatemaximum burial temperatures corresponding to the totalannealing temperature for ZFT. Consequently, in order todecipher the Cenozoic exhumation patterns in the centralsector of the mountain range, we conducted AFT analysis of29 samples, mostly from sedimentary rocks ranging in agefrom Jurassic to Paleocene. AFT ages were obtained byApatite to Zircon, Inc., using the LA-ICP-MS method(see Appendix A). In addition, we present ZFT data from17 samples collected in the Villeta Anticlinorium and theQuetame Massif from pre-Devonian low-grade meta-are-nites and sedimentary units ranging in age from upperPaleozoic to Lower Cretaceous (Figures 1 and 2). ZFTanalyses were conducted at the University of Potsdam,following the external detector method [Gleadow, 1981].On the basis of individual grain age distributions, wedifferentiate concordant (discordant) ages for samples thatpass (fail) the c2 test [Galbraith, 1981; Green, 1981].Fission track (FT) age errors are reported at 1s. Further

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

6 of 27

TC4004

analytical details for the AFT and ZFT measurements aredescribed in Appendix A.[16] Our data set includes samples that were exhumed

from both above and below the depths of total annealingtemperatures for the AFT and ZFT systems. For each ther-mochronological system, we identify samples that wereabove such an isotherm, and hence resided within thePAZ prior to Cenozoic exhumation, by integrating data on(1) paleoburial temperatures based on vitrinite reflectance,(2) the comparison between cooling and depositional ages,(3) the statistical distribution of individual grain ages (con-cordant versus discordant ages according to the c2 test [e.g.,Galbraith, 1981; Green, 1981]), and (4) the distribution ofFT ages with respect to burial depth, as indicated by thestratigraphic depth of the samples [e.g., Brandon et al., 1998;O’Sullivan and Wallace, 2002]. Following these criteria, weidentify three types of FT ages.[17] 1. Reset ages result from samples that attained peak

burial temperatures hotter than the total annealing isothermbefore the last cooling event; reset cooling ages are youngerthan depositional ages and, depending on the homogeneityin the annealing properties of the mineral, may be concor-dant [P(c2) > 5%] or discordant [P(c2) <5%].[18] 2. Partially reset ages occur in samples with peak

burial temperatures similar to or cooler than the total anneal-ing temperature, thus indicating exhumation from within thefission track PAZ. Partially reset cooling ages may overlapwith depositional age and may be concordant or discordant.Partially reset ages have no direct geological meaning sincethey reflect the penultimate cooling event modified by partialannealing.[19] 3. Finally, detrital ages are obtained from samples

with burial temperatures similar to or less than the lowertemperature limit of the PAZ and thus preserved an inheritedage from the source area. Detrital ages are distinguished byhaving a cooling age significantly older than the deposi-tional age and are usually discordant.[20] In order to extract cooling histories from AFT

partially reset and detrital samples, we modeled AFT ages,track lengths and kinetic data following the annealing kinet-ics model of Ketcham et al. [2007b] implemented in theprogram HeFTy [Ketcham, 2005]. We present thermal mod-els for three Paleocene and one Upper Cretaceous samplesfrom the moderately exhumed areas in the axial sector of therange. In addition, we include one partially reset UpperCretaceous sample presented byGomez [2001] in our thermalhistory modeling. Finally, we also modeled AFT data for tworeset samples with available peak burial temperature esti-mates. For these reset samples, thermal modeling extrapo-lates back in time the cooling patterns allowed by the AFTdata in the younger (cooler) portion of the t-T path, in order toinvestigate the approximate age for initiation of cooling.Further analytical details for AFT thermal modeling aredescribed in detail in Appendix B.

5. Thermochronology

[21] In order to provide a more regionally meaningfulevaluation of the exhumation patterns, we integrate our new

data with 24 published AFT and 6 published ZFT ages fromareas to the east of the Central Cordillera at a similar latitude[Gomez, 2001; Gomez et al., 2003;Mora et al., 2008; Parraet al., 2009]. For clarity, we subdivided this thermochrono-logical data set into four transects: the western, northern,central, and southern transects (Figure 1). Raw data andmethodology details for the published AFT analyses arepresented in Table S1. Also, available vitrinite reflectance(R0) data used to constrain peak burial temperatures areshown in Table S2. In this section we present the results andinterpretations according to the different morphotectonicprovinces, moving eastward from the Central Cordilleraand Middle Magdalena Valley basin to the VilletaAnticlinorium, the axial Eastern Cordillera, and finally theQuetame Massif.

5.1. Central Cordillera and Magdalena Valley Basin

[22] Thermochronological data from the Central Cordil-lera at the latitude of our transect are sparse. Gomez et al.[2003] reported a ZFT age of 77.6 ± 3.9 Ma and an AFT ageof 32.0 ± 3.1 Ma from a granodioritic pluton from theeastern Central Cordillera (samples Z01, A01; Figure 3a andTable S1).[23] Farther to the east, in the Eastern Cordillera’s west-

ern fold-and-thrust belt, the timing for the onset of coolingis evaluated through our own thermal modeling of AFT dataobtained by Gomez [2001] and Gomez et al. [2003]. Theypresented AFT ages of 66.2 ± 7.7 Ma and 61.5 ± 9.1 Ma forUpper Cretaceous and lower Paleocene sandstones in thewestern flank of the Guaduas syncline (samples A02–A03,respectively; Figure 3a). Indistinguishable cooling and de-positional ages, a R0 value of �0.6% (�100�C [Burnhamand Sweeney, 1989]) in the Upper Cretaceous BuscavidasFormation [Gomez, 2001] (see Table S3), and a short meantrack length (MTL = 13.56 ± 1.40 mm; Figure 3b) suggeststhat AFT ages are partially reset. Along the same structure,samples A04 and A05, collected �1 to 4 km shallowerwithin the stratigraphic section from upper Paleocene andOligocene sandstones and at higher elevations, yield cool-ing ages of 25.8 ± 3.8 Ma and 12.2 ± 2.3 Ma, respectively(Table S1). Although AFT ages are younger than deposi-tional ages, a virtually invariable R0 value of �0.6%throughout the Upper Cretaceous-lower Miocene sectionin the Guaduas syncline [Gomez, 2001] (Table S3), suggeststhat ages are partially reset.[24] On the basis of thermal modeling of Upper Creta-

ceous and Paleocene samples, Gomez [2001] inferred that afirst cooling episode from peak temperatures of 100�–120�C occurred sometime between 65 and 30 Ma. Also,this cooling event is interpreted to correlate with develop-ment of the regionally extensive angular unconformity atthe base of middle Eocene deposits along the westernfoothills of the Eastern Cordillera and the Middle Magda-lena Valley Basin [Gomez, 2001; Restrepo-Pace et al.,2004]. To better document the timing for such a coolingepisode, we performed thermal modeling on data from threepartially reset sandstones ranging in age from Maastrichtianto late Paleocene (samples A02, A03 and A04, Table S1).Modeling input parameters include AFT grain ages, track

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

7 of 27

TC4004

lengths, and Dpar measurements, as well as R0 values(see Appendix B for details). For samples with a discordantage, we divided the grains into populations on the basis ofAFT kinetic properties (Dpar or Cl content), and modeled thepopulation with the maximum number of grains that yield aconcordant age. We also guided the t-T paths by introducingconstraints related to the time of deposition and, importantly,cooling sometime between 56 and 43 Ma, corresponding tothe time gap associated with the angular unconformity[Gomez et al., 2005]. Only one sample (A02; Maastrichtian

Cimarrona Formation [Gomez et al., 2003]) has a fair numberof confined tracks (50) sufficient for modeling with moderateresolution (Figure 3b). However, the other two samples withfewer tracks yielded similar results (not shown). Modelingresults suggest rapid cooling between peak temperatures anddeposition, implying either a source area rapidly exhumingduring the Late Cretaceous such as the Central Cordillera[e.g., Gomez, 2001; D. Villagomez et al., presented paper,2008], or potential contribution from volcanic sources.Although contemporaneous volcanism is not reported in

Figure 3. Topography, simplified structure, and fission track data from the Central Cordillera and thewestern sector of the Eastern Cordillera. (a) (top) A 40-km-wide topographic swath (SRTM) profileshowing maximum, mean, and minimum elevation, as well as projected FT samples for the Westerntransect (see Figure 1 for location). Vertical dotted lines indicate location of major morphotectonic limits.(bottom) Cooling ages (squares) and depositional ages (stars) for ZFT (open symbols) and AFT (solidsymbols) with bars indicating 1s error. Note cooling ages significantly younger than stratigraphic ages insamples from the Villeta Anticlinorium, reflecting resetting, EC, Eastern Cordillera. (b) Results ofthermal modeling of AFT and R0 data for Upper Cretaceous partially reset sample A02. Good fit t-Tsolutions are shown in dark gray and acceptable fit paths in light gray. The thick dotted path representsthe best t-T solution. The vertical black dotted line indicates the oldest modeled track. Time-temperaturepaths suggest that initiation of cooling occurred sometime between �50 and 35 Ma. GOF is the goodnessof fit between measured and modeled data [Ketcham, 2005].

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

8 of 27

TC4004

Central Colombia, sparse layers of felsic tuff have beenfound interbedded within coal-bearing Paleogene depositsfarther to the north, at �10�N [Bayona et al., 2007]. There-fore, a distal volcanic source may have provided volcanicminerals to nontuffaceous Paleocene foreland deposits in thepresent Eastern Cordillera. Finally, the modeling resultssuggest the time for onset of cooling is constrained tosometime between 50 and 35 Ma.

5.2. Villeta Anticlinorium

[25] Three ZFT samples were collected from LowerCretaceous sedimentary units along the most deeplyexhumed areas in the Villeta Anticlinorium (samples Z02,Z03, Z04; Figure 3a). All samples have concordant agesranging from 19.4 ± 1.1 to 24.0 ± 1.4 Ma (Table 2). A R0

value of 6.4% in Lower Cretaceous rocks nearby indicates apaleoburial temperature > 250� [Gomez, 2001]. Concordantages and peak burial temperature estimates thus indicatereset ZFT ages and imply that exhumation was ongoing by25 Ma. AFT data from equivalent horizons yield reset agesof 6.6 ± 1.8 and 7.9 ± 3 Ma (samples A06 and A07 [Gomez,2001]; Table S1).

5.3. Axial Eastern Cordillera

[26] Along the Northern transect, concordant AFT agesfrom Jurassic and Cretaceous sandstones of the hangingwall block of the Soapaga Fault are significantly youngerthan depositional ages (9.6 ± 2.5 to 25.9 ± 2.2 Ma; Table 1and Figure 4a). This result and R0 values of 0.7–1.0% in theCretaceous units (Guadalupe and Une formations, respec-tively; Table S3) show that ages are reset, and thus exhu-mation must have been active by �26 Ma. A similar coolingversus depositional age relation is observed in the ZFT dataobtained from a Jurassic sandstone and from Precambriangranite and gneiss beneath the Mesozoic sequence (ZFTages = 68.2 ± 7.0 to 89.2 ± 13 Ma; Table 2). However, theseZFT ages are discordant and show large grain age dispersion(42–105%). We thus interpret the ZFT ages as partiallyreset. Similar Cretaceous, partially reset ZFT ages have beenreported in the Santander Massif, 100–150 km to the north[Shagam et al., 1984], reflecting insufficient burial for totalannealing of zircon fission tracks after Jurassic deposition inboth the hanging walls and footwalls of the former trans-tensional Santa-Marta-Bucaramanga fault. In order to in-vestigate the approximate time for onset of cooling in theFloresta Massif, we performed thermal modeling in resetsample A50 from in the Lower Cretaceous Une Formation,combining AFT and R0 data (R0 = �1.0%). In this particularcase, while AFT data constrain the cooling patterns (i.e.,allowed cooling rates) in the range of temperatures at whichfission tracks are preserved in the younger portion of the t-Tpath (<120�C), peak burial temperatures based on R0 data(�150�C) allow extrapolation of this pattern into the older,initial cooling history. Good fit t-T paths suggest a fairlyconstant cooling rate of �3.0–3.5�C/Ma from the �110�Cisotherm since �27 Ma (Figure 4b). Thermal modelingextrapolates this rate into the previous cooling history andhence constrains the initiation of cooling to sometime

between 50 and 30 Ma in the hanging wall block of thecontractionally reactivated Soapaga fault.[27] Farther to the south along the axial sector of the

Eastern Cordillera, a low degree of exhumation has resultedin preservation of an important portion of the Mesozoic, rift-related and Cenozoic foreland sedimentary sequences in thearea of the Bogota Plateau. Thus, along our Central transect,exhumation is evaluated only through AFT dating andthermal modeling. Five sandstone samples from the latePaleocene-early Eocene Lower and Upper Socha formations(samples A08, A09, A13, A14, and A19) yield AFT agesranging from 73.3 ± 3.6 to 38.0 ± 3.5 Ma (Table 1). All butone of the ages (sample A9) are discordant and all are closeto or older than depositional ages (Figure 5a). In addition,R0 values of 0.43–0.52% from these units [Mayorga andVargas, 1995] correspond to peak maximum temperaturesof 80–95�C. Combined, this information indicates thatsamples were exhumed from depths above the total anneal-ing isotherm and thus ages are partially reset. In all but oneof these samples (A08) wemeasured sufficient confined tracklengths for thermal modeling. Modeling was conducted bycombining AFT grain ages, track lengths, and Dpar measure-ments, as well as vitrinite reflectance values (see Appendix Bfor details). In a similar style as thermal models from theMagdalena Valley basin, we used model constraints based onthe known geological history of the region. Particularly, fourconstraints were used for each model: (1) (40�C–200�C)between 100 and 60 Ma, to allow a wide range of predeposi-tional thermal histories, including residence both above andbelow the total annealing isotherm in the source areas of theCentral Cordillera and the Guyana craton; (2) (10�–25�C)during deposition (late Paleocene-early Oligocene, 59–49 Ma); (3) (60�–130�C) between deposition and �15 Ma,to allow heating due to burial reaching temperatures beyondpeak burial temperature estimates; and (4) 20�C at presentconditions. Acceptable and good fits for all models show asimilar t-T path (Figure 5b) that involves three stages. Asthese detrital apatites may be derived from multiple sourceareas, the first modeled cooling cannot be definitively inter-preted. Nevertheless, the first, very rapid cooling overlaps intime with the depositional age, indicating a potential volcaniccontribution from the source area, which has been shown tocorrespond to the Central Cordillera [Gomez et al., 2003].Besides this rapid cooling, the oldest track given for the bestfit path in each model yield Campanian to Paleocene ages,suggesting a potential contribution of apatites from theuplifting Central Cordillera. Second, heating up to 90–105�C precedes the onset of cooling at some time between45 and 30 Ma. Finally, the samples cooled to surface temper-atures at moderate rates of 2�–5�C/Ma.

5.4. Quetame Massif

[28] We present 14 new AFT ages and 11 new ZFT agesfrom samples collected along the Central and Southerntransects in the eastern margin of the Eastern Cordillera(Figure 1 and Tables 1 and 2). Most of the samples werecollected in the hanging walls of the Servita-Lengupa faults(Figures 5 and 6). In aW to E direction the data set includes 5AFT samples from Upper Cretaceous sandstones (Chipaque

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

9 of 27

TC4004

Formation and Guadalupe Group), 9 AFTand 4 ZFTsamplesfrom Lower Cretaceous sandstone units, 1 ZFT sample fromupper Paleozoic sandstones, and 4 ZFT samples from Pre-Devonian low-grade meta-arenites. Finally, 2 ZFT sampleswere collected from the basal Cretaceous sandstones in thefootwall of the Servita fault (Figures 1 and 2).[29] A first-order analysis is derived from comparison of

ZFT ages from equivalent units in different tectonic blocks(Figure 6a). First, two samples from basal Cretaceoussandstones in the hanging wall of the Servita fault havereset Miocene ages of 13.1 ± 0.9 and 9.8 ± 0.6 Ma (samplesZ08 and Z18). Second, another group of samples from the

footwall of this fault yields discordant, partially reset agessimilar to slightly older than the depositional age (samplesZ19 and Z20; 145.2 ± 17.3 and 165.9 ± 12.9 Ma, respec-tively). These results indicate that Lower Cretaceous rockswere buried at depths beneath the ZFT total annealingisotherm only in the western, former hanging wall blockof the extensional Servita fault that formerly defined theeastern margin of the Mesozoic rift. Conversely, as sug-gested by Mora et al. [2008], nearly identical AFT ages of�3 to 4 Ma in Lower Cretaceous and basement samples inthe hanging wall (e.g., samples A29 and A43) and footwall(e.g., samples A45 and A46) blocks of the Servita fault,

Figure 4. Topography, simplified structure, and fission track data from the Floresta Massif area in theaxial sector of the Eastern Cordillera. Symbols and abbreviations are as in Figure 3. (a) (top) A 40-km-wide topographic swath (SRTM) profile showing maximum, mean, and minimum elevation, as well asprojected FT samples for the Northern transect; see Figure 1 for location. (bottom) Cooling ages (1serror) and depositional ages. Note Oligocene-Miocene reset AFT ages from Precambrian to Cretaceousunits. Cretaceous ZFT ages are interpreted as partially reset (see text for discussion). (b) Results ofthermal modeling of AFT and R0 data for Upper Cretaceous partially reset sample A09. Time-temperaturepaths suggest that initiation of cooling occurred sometime between �50 and 30 Ma. The t-T constraint atthe end of the Cretaceous (�70–65 Ma) represents burial during the Late Cretaceous postrift stage. Seetext for further discussion.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

10 of 27

TC4004

Table

1.ApatiteFissionTrack

Data

IDSam

ple

Long(W

)Lat

(N)

Elevation(m

)Unit

Stratigraphic

Age

(Ma)

Number

of

Grains

NSa

S(PW)b

(�10�5cm

2)

1sS(PW)

(�10�7cm

2)

x MSc

1sx M

S

NorthernTransect

A11

RW-2

72�49.0840

5�35.1670

3096

ChipaqueForm

ation

88±5

30

55

2.21

1.51

11.75

0.33

A45

RG

05

73�9.7190

5�54.6300

2600

PalermoForm

ation

188±12

13

40

2.63

6.73

13.84

0.49

A46

RG

04

73�5.2940

5�55.5010

3574

Montebel

Form

ation

168±7

38

77

3.63

1.35

15.85

0.39

A47

RG

01

73�4.2550

5�50.9990

2911

RusiaForm

ation

154±8

14

25

1.89

1.25

16.01

0.41

A48

AM

12

72�53.7170

5�49.4460

2541

Busbanza

Gneiss

>542

31

29

1.46

0.79

16.13

0.42

A49

AM

10

72�52.0610

5�49.9480

2542

GironForm

ation

161±15

33

108

4.45

2.34

16.34

0.43

A50

AM

09

72�51.6080

5�49.6990

2525

UneForm

ation

103±9

30

162

5.21

2.34

16.58

0.45

A51

RW

3b

72�50.2930

5�35.3480

3330

GuadalupeGroup

68±3

18

60

3.66

3.49

11.64

0.32

CentralTransect

A08

RG-06

73�20.9070

5�36.6500

3121

Bogota

Form

ation

54±5

18

291

4.27

6.65

13.77

0.48

A09

M4-04

73�21.8740

5�22.0840

2147

Lower

SochaForm

ation

59±3

40

614

4.78

3.59

11.47

0.30

A10

T4-032

73�11.0080

5�18.4280

1815

UneForm

ation

103±9

38

51

4.42

2.54

11.37

0.29

A12

AM-01

73�26.2730

5�13.1120

2097

PicachoForm

ation

46±2

13

51

3.81

5.06

17.49

0.52

A13

AM-03

73�25.5170

5�13.2440

2025

Upper

SochaForm

ation

54±5

37

134

3.03

1.80

17.38

0.51

A14

AM-04

73�24.4720

5�13.8080

1901

Upper

SochaForm

ation

54±5

37

297

4.84

2.11

17.18

0.50

A15

AM-05

73�23.7230

5�13.2600

1928

GuadalupeGroup

68±3

25

63

2.52

1.01

17.04

0.49

A16

AM-06

73�22.9210

5�11.5690

1711

UneForm

ation

103±9

35

14

2.22

1.06

16.84

0.47

A17

T3-004

73�22.4630

5�10.7540

1884

UneForm

ation

103±9

24

62

4.55

2.93

15.27

0.35

A18

T3-003

73�20.5620

5�11.0390

2396

UneForm

ation

103±9

12

13

0.98

1.11

15.38

0.36

A19

RG-16

73�45.4930

5�16.2690

2595

Upper

SochaForm

ation

54±5

39

595

10.11

4.95

15.65

0.38

A20

RG-14

73�45.9580

5�15.5000

2600

CachoForm

ation

59±3

14

83

3.23

6.51

13.68

0.46

A21

T2-018

73�42.8970

5�6.4720

2630

GuadalupeGroup

68±3

868

1.85

4.55

13.31

0.39

A22

T2-660

73�37.7620

5�3.1100

2275

UneForm

ation

103±9

29

16

0.85

0.67

15.52

0.37

A23

T2-013

73�37.5630

5�2.5450

2255

UneForm

ation

103±9

21

55

3.80

7.31

13.44

0.42

A24

T2-012

73�36.6020

5�0.7940

2687

GuadalupeGroup

77±6

20

232

6.50

10.11

13.56

0.44

A25

T2-019

73�32.9760

5�0.1580

1868

UneForm

ation

103±9

53

0.37

0.70

13.23

0.38

A26

AM-07

73�23.4410

5�1.9250

1341

Las

JuntasForm

ation

117±3

118

3.58

3.14

16.68

0.46

SouthernTransect

A35

BV-M

P3-F

74�1.2390

4�26.5610

2275

ChipaqueForm

ation

85±5

18

28

1.13

1.37

11.14

0.27

A36

BV-M

P7-F

74�0.5610

4�26.0940

2022

ChipaqueForm

ation

85±5

10

14

0.54

0.91

11.07

0.26

A37

BV

120-F

73�57.2560

4�23.8720

2123

CaquezaForm

ation

133±2

39

82.24

3.04

11.22

0.28

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

11 of 27

TC4004

Table

1.(continued)

ID

43Cad

(�10�2)

238Ue

(�10�2)

P(c

2)f

(%)

Age

(Ma)

±1sError

Dpar

(mm)

SD

(mm)

Number

ofDpar

Measurements

Length

(mm)

Error

(mm)

SD

(mm)

Number

ofLength

Measurements

NorthernTransect

A11

3.89

10.08

0.0

14.1

2.0

1.62

0.17

30

13.71

0.54

1.52

9A45

3.81

0.94

7.5

10.5

1.7

1.50

0.27

12

12.73

1.22

1.73

3A46

3.48

4.63

76.5

19.9

2.1

1.59

0.15

37

13.61

0.22

1.6

53

A47

3.73

9.29

15.5

10.6

2.1

1.57

0.26

13

14.12

0.43

1.23

9A48

2.73

4.32

72.8

16.0

3.0

1.59

0.15

30

13.89

0.21

1.55

55

A49

3.78

23.08

3.4

19.8

2.0

1.66

0.23

32

14.22

0.19

1.42

54

A50

2.68

4.78

5.6

25.9

2.2

1.59

0.20

74

12.91

0.21

1.84

75

A51

4.77

4.60

0.0

9.6

2.5

1.53

0.22

17

12.66

0.62

1.64

8

CentralTransect

A08

3.53

0.49

0.0

46.8

3.3

1.51

0.13

17

13.50

0.82

2.58

11

A09

3.42

0.82

11.3

73.2

3.6

2.60

0.52

40

14.18

0.1

1.47

203

A10

4.12

1.86

97.9

6.6

0.9

1.71

0.39

38

13.00

0.48

2.05

19

A12

2.72

8.32

0.0

11.7

1.7

1.59

0.22

12

12.90

0.53

2.05

16

A13

1.43

6.39

0.0

38.3

3.5

2.15

0.49

35

13.53

0.14

1.79

156

A14

1.97

2.13

2.5

52.5

3.4

2.55

0.47

34

13.19

0.16

2.05

163

A15

3.67

16.43

0.0

21.2

2.7

1.68

0.3

24

13.03

0.37

2.05

32

A16

2.35

5.44

0.0

5.3

1.4

1.54

0.16

28

12.06

0.66

1.63

7A17

2.96

1.34

45.3

10.4

1.3

1.60

0.2

23

11.96

0.35

1.91

30

A18

2.47

4.75

95.0

10.2

2.8

1.69

0.19

11

9.67

1.21

3.21

8A19

2.68

3.40

0.3

45.9

2.2

1.88

0.23

35

12.74

0.17

2.06

156

A20

4.14

0.70

4.6

17.6

2.0

1.54

0.27

13

13.30

2.42

4.19

4A21

4.03

75.87

0.0

24.4

3.1

1.57

0.25

713.45

3.8

3.8

2A22

2.93

19.48

0.0

14.6

3.7

1.61

0.21

14

14.04

0.33

0.8

7A23

3.49

4.20

27.3

9.7

1.4

1.51

0.19

19

10.37

1.41

3.16

6A24

4.20

0.34

0.0

24.2

1.8

1.59

0.19

20

12.97

0.6

2.69

21

A25

3.88

5.64

66.8

5.4

3.1

1.64

0.15

5-

--

-A26

2.92

6.81

57.0

1.9

0.7

1.52

0.23

11

12.10

0.29

0.77

8

SouthernTransect

A35

4.02

7.21

23.8

13.8

2.6

1.64

0.13

17

12.35

1.64

3.27

5A36

2.86

2.63

0.1

14.3

3.8

1.68

0.25

912.28

0.51

0.72

3A37

4.03

49.22

99.9

2.0

0.7

1.62

0.22

28

13.06

0.95

0.95

2

aNumber

ofspontaneousfissiontrackscountedover

area

W.

bSum

ofPi*Wiforallgrainsevaluated;Piis(238U/43Ca)

forapatitegrain

i;Wiisarea

over

whichNSandPiareevaluated.

cThezcalibrationfactorbased

onLA-ICP-M

Soffissiontrackagestandards.

dBackground-corrected

43Ca(dim

ensionless).

eBackground-corrected

238U

(dim

ensionless).

f Chi-squareprobability.

Values

greater

than

5%

areconsidered

topassthistestandrepresentasingle

populationofages.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

12 of 27

TC4004

Table

2.ZirconFissionTrack

Data

IDSam

ple

Long

(W)

Lat

(N)

Elevation

(m)

Unit

Strat.Age

(Ma)

Number

ofGrains

U(ppm)

Rho-S

(NS)a

Rho-I

(NI)a

Rho-D

(ND)b

P(c

2)c

(%)

Age

Dispersion

Aged

±1sError

Western

Transect

Z02

Villeta

74�24.7240

5�7.0960

737

Villeta

Group

135±5

26

84

10.597(1406)

12.564(1667)

4.6803(3924)

8.7

11%

24.0

1.4

Z03

Murca1

74�20.6070

5�15.2570

1114

MurcaForm

ation

138±2

21

224

21.759(2252)

29.817(3086)

4.3912(4082)

14.7

6%

19.5

1.0

Z04

Murca2

74�19.7150

5�14.6750

1085

MurcaForm

ation

138±2

20

248

27.739(1769)

38.244(2439)

4.4023(4082)

11.0

9%

19.4

1.1

NorthernTransect

Z21

FS17

73�0.1900

5�53.6060

2975

GironGroup

161±15

25

330

112.508(2846)

46.529(1177)

4.597(2370)

0.0

42%

68.2

7.0

Z22

FS15

72�53.8120

5�49.0100

2515

Gneiss

>542

21

209

110.987(2333)

30.732(646)

4.6245(2370)

0.0

105%

68.3

16.4

Z23

FS11A

72�51.8690

5�50.9250

2789

Granite

>542

13

270

141.762(1229)

41.064(356)

4.6521(2370)

0.0

44%

89.2

13.0

CentralTransect

Z05

Soescol

73�23.0650

4�51.2050

1858

Macanal

Form

ation

135±5

8474

45.929(240)

77.697(406)

4.3763(4082)

90.8

0%

15.7

1.5

Z07

MA

16

73�16.0350

4�52.3930

990

Farallones

Group

388±28

35

219

26.236(2354)

28.543(2561)

4.4485(2370)

9.5

10%

24.9

1.4

SouthernTransect

Z08

FT1

73�43.8700

4�29.3080

3660

ChingazaForm

ation

143±2

46

402

26.654(2401)

55.351(4986)

4.4272(3924)

0.0

28%

13.1

0.9

Z09

SJ5

73�41.5860

4�28.2960

2251

Quetam

eGroup

>416

43

322

19.038(1473)

47.097(3644)

4.6695(3924)

88.1

0%

11.5

0.6

Z10

SJ1

73�40.8510

4�29.6280

2748

Quetam

eGroup

>416

14

544

51.06(1519)

16.942(504)

2.9134(2766)

6.2

15%

5.9

0.4

Z13

BV

126

73�54.5230

4�22.6290

1582

Macanal

Form

ation

135±5

11

322

22.887(458)

37.379(748)

3.5714(3388)

5.1

16%

13.3

1.0

Z15

BV

86

73�50.0750

4�16.7320

1226

Quetam

eGroup

>416

21

350

13.470(927)

35.28(2428)

3.5558(3091)

23.2

5%

8.3

0.5

Z17

BV

196

73�47.0020

4�17.5500

2862

Quetam

eGroup

>416

38

420

26.636(1705)

57.880(3705)

4.3073(3924)

0.0

50%

11.4

1.1

Z18

BV

194

73�46.6280

4�17.8550

3084

ChingazaForm

ation

143±2

27

297

14.885(1058)

44.937(3194)

4.8631(3924)

61.4

3%

9.8

0.6

Z19

BV

283

73�36.5450

4�16.7730

1096

Buenavista

Form

ation

144±2

20

231

187.07(3799)

26.344(535)

3.5478(2766)

0.0

44%

145.2

17.3

Z20

BV

279

73�40.1410

4�9.7430

870

Buenavista

Form

ation

144±2

20

204

185.783(6244)

235.65(792)

3.6288(2766)

0.0

21%

165.9

12.9

aRhoSandRhoIarethespontaneousandinducedtracksdensity

measured,respectively(�

105tracks/cm

2).NSandNIarethenumber

ofspontaneousandinducedtrackscountedforestimatingRhoSand

RhoI,respectively.

b(c

2)(%

)isthechi-squareprobability[G

albraith,1981;Green,1981].Values

greater

than

5%

areconsidered

topassthistestandrepresentasingle

populationofages.

cRhoDistheinducedtrackdensity

measuredin

theexternalmicadetectorattached

toCN2dosimetry

glass

(�105tracks/cm

2).NDisthenumber

ofinducedtrackscountedin

themicaforestimatingRhoD.

dPooled(central)ageforsamplespassing(failing)thec2test.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

13 of 27

TC4004

Figure

5

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

14 of 27

TC4004

suggest that Pliocene exhumation was accompanied bythrusting along a single block limited to the east by theMirador short-cut fault (Figure 7).[30] On the other hand, AFT reset ages from the whole set

of Pre-Devonian to Upper Cretaceous samples are younger

than 25 Ma and hence reveal ongoing exhumation byOligocene time. In order to evaluate the time for the onsetof exhumation, provided by the structurally shallowestsample beneath the lower boundaries of the now exhumedAFT and ZFT partial annealing zones, we analyze the

Figure 6. Topography, simplified structure and fission track data from the axial sector Eastern Cordilleraalong the Southern transect (see Figure 1 for location). Symbols and abbreviations are as in Figure 3. (a) A40-km-wide swath-averaged maximum, mean, and minimum topographic (SRTM) profiles. Projected FTsamples are shown. (b) Cooling ages (1s error) and depositional ages. Note overlapping depositional andZFT ages in the footwall of the Servita fault, suggesting partial resetting (see text for discussion). Mioceneexhumation of the Quetame Massif is illustrated by ZFT and AFT ages younger than 20 Ma.

Figure 5. Topography, simplified structure, and fission track data from the axial sector Eastern Cordillera. Symbols andabbreviations are as in Figure 3. (a) (top) A 40-km-wide swath-averaged maximum, mean, and minimum topographic(SRTM) profiles for the Central transect (see Figure 1 for location). Projected FT samples are shown. (bottom) Cooling ages(1s error) and depositional ages. Note similarity between AFT ages and depositional ages for the Paleocene and youngersamples in the western half of the transect, indicating that ages are partially reset. (b) Results of thermal modeling of AFTand R0 data for and Paleocene samples A13, A14, and A19. The portion of the thermal histories older than the oldestmodeled track (indicated by vertical line in each model) is not resolved by our thermal input data; however, they areconsistent with known geological constraints.

15 of 27

TC4004TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

distribution of fission track ages with respect to stratigraphicdepth [e.g., Brandon et al., 1998; O’Sullivan and Wallace,2002] for samples collected from the hanging wall block ofthe Servita and Lengupa faults in the Central and Southerntransects, including both our new ages and published data(see Table S3). This approach constitutes an approximationto the age-elevation profiles that have been used to derivethe timing and rates of exhumation in exhumed crustalsections [e.g., Fitzgerald et al., 1993; Ehlers, 2005] or thepaleodepth-age profiles used in extensional settings [Stockliet al., 2002; Stockli, 2005]. Figures 8a and 8b show thedistribution of AFT and ZFT ages with respect to thestratigraphic datum defined by the angular unconformityat the base of the Lower Cretaceous rift deposits (seeTable S3). On this plot, the distribution of AFT ages in theCentral transect (Figure 8a) defines a major break in slope

between the top of the Upper Cretaceous Guadalupe Groupand the late Paleocene Lower Socha Formation (between 6000and 6500 m above the unconformity), with an age between�25 and�40Ma. R0 values of 0.6–0.7% at this stratigraphicdepth suggest peak burial temperatures of 110–120�C [afterBurnham and Sweeney, 1989]. This temperature rangecorresponds to the total annealing temperature for moder-ately rapidly cooled, fast annealing apatites, typically char-acterized by Dpar values <1.75 mm [Donelick et al., 2005],such as those observed in the Cretaceous samples along thistransect (Figure 8a, right). Collectively, the stratigraphicdepth-age distribution for the kinetically homogeneousapatites and the R0 data thus suggest that the observedbreak in slope corresponds to an approximate location of thebase of an exhumed AFT PAZ. However, since our analysiscombines samples from different structural levels over a

Figure 7. (a) AFT and ZFT data collected along the Southern transect. The easternmost ZFT sample(Z20) collected from Lower Cretaceous sedimentary units in the footwall of the Servita thrust (ST) yield aCretaceous age, indicating a provenance age modified by partial annealing. ZFT samples from thehanging wall of the thrust display a pattern of increasing ages toward shallower structural levels. The baseof the partial annealing zone for ZFT is located between samples Z12 and Z13. See Figure 8 and text forfurther discussion. (b) Structural cross section after Mora et al. [2008].

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

16 of 27

TC4004

Figure 8

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

17 of 27

TC4004

horizontal distance of �70 km, we use such an approximatelocation of the AFT PAZ to derive a conservative estimateof the time of onset of cooling. Our data suggest that coolingwould have commenced sometime between 25 and 40 Ma.Furthermore, more resistant apatites (with larger Dparvalues) in the overlying Paleocene samples (�6500–7000m) have older ages, supporting our interpreted position forthe base of the PAZ. Sample A12, obtained from theyoungest unit in the profile, has an anomalously youngage (11.7 ± 1.7 Ma) compared to nearby samples. Wesuspect this may arise from sample A12 residing in adistinct faulted unit not recognized in the field.[31] A similar analysis was conducted for ZFT samples

collected along the Southern transect (Figure 8b). Here, amajor break in slope of a stratigraphic depth-age plot occurswithin the Macanal Formation, between 1500 and 2200 mabove the base of the Cretaceous units (Figure 8b). R0

values > 3.5% for the Macanal Formation [e.g., Toro et al.,2004; Mora et al., 2008] suggest peak temperatures similarto that of total annealing for ZFT, and hence reveal the baseof an exhumed zircon fission track PAZ. In this transect, thestructural configuration of the sampled section allows acloser approximation of the time for the initiation of cool-ing. As shown in Figure 7, we identify the position of thelower boundary of the PAZ based on the ages obtained insandstones from the Macanal Formation cropping out in thecore of the Caqueza anticline within the hanging wall blockof the Naranjal fault. Here, samples Z11 and Z12 yielddiscordant ages of 136.5 ± 13.4 Ma and 61.3 ± 3.4 Ma,respectively; we interpret these ages as partially reset. Incontrast, structurally deeper samples Z13 and Z14 yieldconcordant reset ages of 13.3 ± 1.0 Ma and 18.5 ± 1.0 Ma.Therefore, according to the pattern of reset ages from thelower part of the Macanal Formation and the underlyingbasement units, the age for onset of cooling must be veryclose to that of the uppermost totally annealed sample (Z14;18.5 ± 1.0 Ma, 1510 m). We thus interpret that thrust-induced exhumation started at �20 Ma in the hanging wallblock of the Naranjal fault.[32] To better constrain the timing and apparent rates of

exhumation, we constructed stacked pseudovertical profiles[e.g., Reiners et al., 2003] combining AFT and ZFT data forthe Central and Southern transects (Figures 8c and 8d).While AFT ages are plotted in their actual stratigraphicposition, elevations for ZFT data are shifted upward by anamount proportional to the difference between the depths ofthe total annealing isotherms for both thermochronometricsystems. Using typical thermal parameters for convergent

orogens within a stable thermal regime and moderate exhu-mation rates of 0.1–0.3 mm/a, this difference has beenestimated to be �5.7 km [Reiners and Brandon, 2006]. Inthe Eastern Cordillera of Colombia, however, acceleratedPliocene exhumation, as suggested by AFT data [Mora et al.,2008] (see Figure 8), implies an unstable thermal regimedue to compression of the isotherms and increase of thegeothermal gradient by heat advection [e.g., Mancktelow andGrasemann, 1997]. Nevertheless, we use this depth differenceof 5.7 km between the ZFTand AFT total annealing isotherms(equivalent to a difference of 115–140�C for geothermalgradients of 20–25�C, similar to the modern value in theproximal foredeep region [Bachu et al., 1995]), since the timewindow covered by our ZFT data largely predates the episodeof Pliocene-Pleistocene accelerated exhumation.[33] Both stacked pseudovertical profiles show similar

exhumation patterns (Figures 8c and 8d) characterized bymoderately rapid exhumation followed by much fastererosion beginning in the last �4–5 Ma. Using the calcu-lated pseudomean thicknesses, ZFT and AFT ages overlapin both profiles, suggesting that the estimated depth differ-ence between both isotherms is correct. A first break inslope at �25–40 Ma in the northern profile and �20 Ma tothe south marks the timing for the onset of thrust-inducedexhumation along the hanging wall of the Servita-Lengupafaults. On the basis of a subset of 8 samples collected alongthe Southern transect (Figure 8d), we derive an error-weightedpseudoapparent exhumation rate of 0.3 ± 0.1 mm/a between�20 and �9 Ma. Although this rate is not directly compara-ble with apparent exhumation rates derived from true age-elevation profiles, we use it as a minimum estimate as itassumes exhumation trajectories perpendicular to bedding,and hence neglects folding accompanying exhumation. Asecond break in slope observed in both transects placesmaximum constraints on the timing for the onset of accel-erated exhumation at �4–5 Ma. As noted by Mora et al.[2008], based on a true AFT vertical profile, this episode ofaccelerated exhumation (1–2 mm/a) was restricted to theeastern, windward flank of the Eastern Cordillera andsuggests a positive feedback between the buildup of topog-raphy, focused orographic precipitation, and long-term ex-humation rates.

6. Discussion

6.1. Diachronous Exhumation

[34] In absence of other mechanisms that could havegenerated regional cooling in the Eastern Cordillera, such

Figure 8. Fission track data and vitrinite reflectance (R0) values for samples from the (a) Central and (b) Southerntransects plotted against the stratigraphic position referred to the base of the Cretaceous rift-related units (see Table S3 fordetails and Figure 1 for location). Stratigraphic thicknesses and ages are compiled from Ulloa and Rodrıguez [1979] andMora et al. [2008]. Vertical green bars represent the range of R0 values that correspond to the temperature delimiting thebase of the AFT (Central transect) and ZFT (Southern transect) partial annealing zones (blue and pink shaded areas,respectively). Stacked pseudovertical profiles are obtained for the (c) Central and (d) Southern transects. AFT data areplotted at their original stratigraphic position, as in Figure 8a and 8b, but ZFT data are offset upward by an amountproportional to the depth difference between the ZFT and the AFT isotherms, estimated to be 5.7 km. The first break inslope, denoted by vertical light gray band, at �40–25 Ma (central profile) and 20 Ma (southern profile) marks the onset ofthrust-induced cooling through the AFT and ZFT total annealing isotherms, respectively. See text for discussion.

18 of 27

TC4004TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

as postmagmatic cooling [e.g., Tagami and Shimada, 1996]or normal faulting driving tectonic exhumation [e.g., Ring etal., 1999], we interpret the cooling ages obtained in thisstudy as a signal of erosional exhumation. Since erosion isfacilitated by the creation of topography and the formationof pronounced relief contrasts accompanying reverse fault-ing in contractional orogens, the age for initial cooling inthe study area constitutes a proxy for initial thrusting.[35] Figure 9 summarizes the spatiotemporal pattern of

onset of thrust-induced exhumation across the easternColombian Andes. To the west, in the Central Cordillera,a ZFT age of �78 Ma suggests that exhumation was activeby late Cretaceous (Campanian) time. Existence of elevatedtopography in the area of the Central Cordillera in the LateCretaceous has also been inferred from paleoflow data,sandstone and gravel petrography, and the character anddirection of facies changes in Upper Cretaceous conglom-erate and sandstone units in the Guaduas syncline [Gomez etal., 2003], and farther to the south, in the upper MagdalenaValley basin [Montes et al., 2005; Ramon and Rosero,2006]. On the other hand, an AFT age of �32 Ma [Gomezet al., 2003] places additional constraints on the Cenozoiccooling patterns in this sector of the Central Cordillera.Assuming total annealing temperatures of 250 ± 40�C and120 ± 10�C for the ZFT and AFT, respectively, a geothermalgradient of 20�–25�C, surface temperatures of 20�C, andlinear cooling, these ages indicate moderate exhumation ratesof 0.1–0.2 mm/a since the late Cretaceous.[36] Eastward advance of the orogenic front is docu-

mented by the estimated age for initial cooling in thewestern fold-and-thrust belt of the Eastern Cordillera. Here,development of an angular unconformity between the latePaleocene Hoyon Formation and the middle Eocene San

Juan de Rio Seco Formation brackets the initial time forrock uplift in the Guaduas syncline to 56–43 Ma [Gomez etal., 2003]. Likewise, farther to the NE, at �6�N, westverging thrusts are cut off beneath the angular unconfor-mity at the base of the middle Eocene [Restrepo-Pace et al.,2004]. Thermal modeling of partially reset AFT ages in anUpper Cretaceous sample from the western flank of theGuaduas syncline suggests that cooling began sometimebetween 50 and 35 Ma. Within this structure, AFT data byGomez [2001] show an inverted pattern of younger AFTages in stratigraphically shallower and topographically higherlevels (see Figure 3a). Along the same section, a similar burialtemperature for the Upper Cretaceous (Buscavida Formation)to Miocene (Santa Teresa Formation) units is suggested bynearly invariable R0 values of �0.6% throughout the entirestratigraphic section [Gomez et al., 2003] (Table S2). Weinterpret this pattern as an indicator of pre-Miocene growthof the western flank of the Guaduas syncline, which wouldhave prevented the stratigraphically deeper portion of thesection from reaching burial beneath the full stratigraphicthickness.[37] Farther to the east, three reset ZFT ages of �24–

19 Ma from Lower Cretaceous units in the core of theVilleta Antilcinorium constrain the minimum age for thrust-driven exhumation associated with contractional reactiva-tion of the La Salina fault. In this area, our data are notsufficient for delimiting the ZFT PAZ, which would requiresampling shallower structural levels with partially resetages. In addition, partially reset AFT age data which couldaid in resolving the patterns of initial cooling throughthermal modeling are not available. However, growth stratawithin the middle Eocene to Oligocene San Juan de RioSeco Formation in the eastern flank of the Guaduas syncline

Figure 9. Summary of chronological indicators of initiation of contractional deformation in the easternColombian Andes. Black bars represent the maximum and minimum age estimates inferred fromindicators of brittle deformation preserved in sedimentary basins across the range. Gray bars representindicators based on thermochronometric data. See text for discussion. Sources are 1, Gomez et al. [2003];2, Gomez [2001]; 3, this work; 4, Restrepo-Pace et al. [2004]; 5, Julivert [1963]; 6, Gomez et al. [2005];and 7, Parra et al. [2009].

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

19 of 27

TC4004

indicate a maximum age of �43 Ma for initial deformationalong the La Salina fault [Gomez et al., 2003]. On the otherhand, AFT cooling ages of 7–10 Ma in the core of theVilleta Anticlinorium reveal Miocene exhumation associat-ed with thrusting of the La Salina fault. Simple calculationsconsidering pooled ZFT and AFT ages, as well as assumingtotal annealing temperatures of 250 ± 40�C and 120 ± 10�Cfor the ZFT and AFT systems, respectively, linear cooling,and horizontal isotherms, suggest that long-term exhuma-tion rates must have increased between twofold and fourfoldduring the late Miocene-Pliocene.[38] In the axial sector of the Eastern Cordillera, our

thermochronologic results from Jurassic and older basementrocks in the Floresta Massif show reset Miocene AFT ages(samples A45–A49; �11–20 Ma) and Cretaceous partiallyreset ZFT ages (samples Z21-23) in all structural levelssampled. These results are strikingly similar to thoseobtained by Shagam et al. [1984] in the Santander Massif,�100–150 km to the north, and demonstrate that Creta-ceous to Cenozoic overburden was significantly thinner inthe northern sector of the range than in the southern sectornear the Quetame Massif, where ZFT ages are reset. In theFloresta Massif, Miocene reset cooling ages in Jurassic andolder rocks and, in particular, an AFT reset age of �26 Main a Lower Cretaceous sandstone from the Une Formation(sample A50) reveal pre-late Oligocene initiation of exhu-mation in the Floresta Massif. Furthermore, this AFT resetage from Lower Cretaceous sandstones in the western,hanging wall block of the Soapaga fault is older thancooling ages retrieved from coeval and younger units tothe east, in its footwall (Une, Chipaque, and Guadalupeformations; samples A10, A11, A51; �6–15 Ma). Thispattern reveals that pre-late Oligocene exhumation wasassociated with motion on the Soapaga fault. Thermalmodeling including estimates of peak burial temperaturesfor the reset AFT sample A50 limits the initiation of thrust-induced exhumation in the western, hanging wall block ofthe Soapaga fault to Eocene-early Oligocene time (�50–30 Ma). Approximately 100 km to the south, along theCentral transect, a similar middle Eocene-early Oligoceneage (�45–30 Ma) for the onset of cooling is determined byAFT thermal modeling of partially reset late Paleocenesedimentary rocks from the Bogota Plateau (Figure 5).These ages for the onset of thrust-induced exhumation inthe Bogota Plateau and Floresta Massif coincide with theformation of middle Eocene-Oligocene growth strata southof the plateau (Regadera and Usme formations [Julivert,1963; Gomez et al., 2005]). In addition, reconstruction ofthe subsidence history in the Medina wedge top basin[Parra et al., 2009] and in the Llanos Basin at �5.5�N[Bayona et al., 2008] suggest a major increase in tectonicsubsidence at around 30–35 Ma (late Eocene-early Oligo-cene). Parra et al. [2009] interpreted this as reflectingeastward advance of the orogenic load and uplift of theQuetame Massif, based on the absence of any other discretetectonic block to the west along a cross section at �4.5�N.However, our extended thermochronologic data set andgrowth strata in the area of the Bogota Plateau support anEocene-late Oligocene initiation of thrust-driven exhuma-

tion along the present-day axis of the Eastern Cordillera. Tothe north, in the Floresta Massif, this exhumation resultedfrom tectonic inversion of a Mesozoic half graben basinlimited to the east by the Soapaga fault. To the south, ourstructural cross section across the Bogota Plateau suggeststhat there is no particular structure that can delimit a tectonicblock associated with this Eocene-late Oligocene exhuma-tion (Figure 1b). We thus infer that contractional deforma-tion in this location was accommodated along a previouslyunrecognized blind fault that would be located southwardalong the structural trend of the Soapaga fault. Takentogether, our new thermochronologic data, the age rangeof synkinematic sediments in the Bogota Plateau, andsubsidence histories in the modern wedge top and foredeepbasins to the east suggest that the Soapaga fault and itssouthern continuation formed the leading edge of deforma-tion in this sector of the northern Andes at 35–30 Ma.[39] Our thermochronologic data do not support a previ-

ous hypothesis that on the basis of the presence of unstablelithic fragments in upper Paleocene and Miocene sandstonesfrom the Eastern Cordillera, erosion of metamorphic base-ment in the Floresta Massif occurred in Paleocene toMiocene time [Bayona et al., 2008]. Such a hypothesisimplies an unrealistic scenario that requires erosional re-moval of �5–6 km of Paleozoic, Jurassic and Cretaceoussedimentary overburden from the area of the present-dayFloresta Massif in the early to middle Paleocene (�7 Ma),followed by negligible unroofing of basement rocks fromlate Paleocene time to the present, i.e., a time span of 56 Ma.Conversely, the Miocene reset AFT ages obtained in thisstudy from Jurassic and basement units in the FlorestaMassif (samples A45–A49; �11–20 Ma) imply that fornormal geothermal gradients of 20–25�C/km, samples fromJurassic and older rocks were buried at a depth of 3 to 4 kmduring the Miocene, at the time of closure. A similarscenario for the igneous and metamorphic basement in theSantander Massif is inferred in light of the similar AFT andZFTage patterns with respect to the Floresta Massif [Shagamet al., 1984]. Considering the �3 km of stratigraphic thick-nesses for the Cretaceous section in the Floresta Massif[Etayo-Serna, 1968] the AFT cooling ages suggest unroof-ing of Cretaceous units during the Miocene. Such a Mioceneerosion window exposing Cretaceous rocks in the axialEastern Cordillera is further supported by the presence ofrecycled detritus of Cretaceous sedimentary rocks in theOligocene-Miocene units of the Eastern Cordillera. SuchCretaceous recycled detritus include gravel clasts of glauco-nitic sandstone in conglomerates [Parra et al., 2007], frag-ments of chert and foraminifera-bearing siltstones inconglomerate and sandstone [Cespedes and Pena, 1995;Bayona et al., 2008], and Cretaceous palynomorphs withinorganic matter-rich siltstones and mudstones [Bayona et al.,2008; C. Jaramillo, personal communication, 2008].[40] Farther eastward propagation of the actively exhum-

ing areas is documented by the pattern of ZFT and AFTreset ages in the Quetame Massif. Age-stratigraphic depthprofiles and the pattern of ZFT ages in the core of theCaqueza anticline, to the west of the Naranjal fault, suggestan onset of exhumation at �20 Ma. Ongoing Miocene

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

20 of 27

TC4004

exhumation associated with this fold is further confirmed byreset AFT ages of �14 Ma in shallower structural levelswithin the Cretaceous section (samples A35–A36). Fartherto the east, we obtained reset ZFT ages as old as �12 Ma inthe structurally shallower levels exposed in the basementhigh. Therefore, our ZFT data cannot directly constrain theonset of cooling in the eastern, hanging wall block of theNaranjal fault. Such information would require employing ahigher-temperature thermochronometric system. However,here we discuss additional evidence based on geodynamicmodeling, sediment accumulation rates in the adjacentbasins, and the structural configuration of the easternmargin of the Eastern Cordillera that suggest uplift of theQuetame Massif beginning in the early Miocene. First, two-dimensional flexural models require tectonic loads in theQuetame Massif to match the reconstructed middle Mioceneflexural profile in the Llanos basin [Bayona et al., 2008].Second, an increase in the sediment accumulation ratesrecorded by the early to late Miocene Upper CarboneraFormation in the Medina wedge top basin [Parra et al.,2007; Gomez et al., 2009; Parra et al., 2009] and the Llanosforedeep [Bayona et al., 2008] suggest an eastward advanceof the orogenic front and the corresponding loads from theirposition along the axial Eastern Cordillera during the lateEocene-Oligocene. Finally, the offset of the basal Creta-ceous unconformity across the Cenozoic faults, as noted byMora et al. [2006, 2009], illustrates that the Naranjal faultis one of many Mesozoic normal faults that have undergoneonly minor Cenozoic contractional reactivation (see Figure 7).Instead, these inherited structures have been passivelyuplifted atop a major basement thrust along the formerlyrift-bounding Servita fault (Figure 1b). This structuralconfiguration suggests that orogenic loading in the QuetameMassif could only be generated by slip along this majorbasement thrust. Taken together, this suggests that contrac-tional reactivation of the Naranjal fault may have beencoeval with slip along the Servita fault. This also corrobo-rates the interpretation that thrust-related cooling of theentire hanging wall block of the Servita fault began atapproximately 20 Ma.

6.2. Pattern of Orogenic Front Advance

[41] The initiation of rock uplift associated with specificstructures, as well as the reconstruction of their retrode-formed palinspastic positions allows a calculation of the rateof migration of the orogenic front toward the foreland overtime. First, the propagation rate of the orogen is the rate ofits lengthening, which is obtained by dividing its presentlength in the transport direction by the time elapsed since theonset of shortening [DeCelles and DeCelles, 2001]. Inthe Colombian Andes, we used the present distance betweenthe eastern boarders of the Central and Eastern Cordilleras(�200 km), and the age of the thrust belt (82–74 Ma) toderive a long-term propagation rate of 2.5–2.7 mm/a.Second, we evaluate the rate of propagation rates betweenadjacent structures in order to resolve the spatiotemporalpattern of orogen growth. The upper and lower limits for theamount of orogen lengthening between such structures isprovided by the restored, and the present (i.e., shortened)

distances between them, respectively (Table 3). We use thebalanced structural cross section from Mora et al. [2009] tocalculate the restored distances between the structures and apin line defined at the surface trace of the most frontalstructure, the Guaicaramo thrust (Figure 10). To the west ofthe western end of the cross section, we use the palinspasticreconstruction of Gomez et al. [2003] to estimate therestored distance between the Central Cordillera’s deforma-tion front and the Cambao thrust. However, limited infor-mation on the deep structure and amount of shorteningassociated with deformation in the Central Cordillera pre-cludes a more precise palinspastic reconstruction of this partof the orogen. A closer look at the spatiotemporal advanceof the orogenic wedge suggests a three-stage evolution(Table 3 and Figure 10). First, after initial uplift of theCentral Cordillera at �82–74 Ma, deformation migrated tothe Cambao fault and associated west verging thrust systemsat a rate of 0.5–2.1 mm/a and continued at similar toslightly higher rates of 0.5–3.1 mm/a during the latePaleocene-middle Eocene as deformation propagated west-ward to the La Salina fault. By that time the orogen hadattained as much as 26% of its present width. Second, avirtually simultaneous uplift along this fault and the axialsector of the Eastern Cordillera along the Soapaga faultduring the middle-late Eocene (30–40 Ma) illustrates arapid eastward jump of the deformation front. The broadage range for the onset of deformation is due to the low-resolution biostratigraphical age determinations of thegrowth strata, resulting in a large uncertainty in the calcu-lation of the rate of migration. However, the data suggestvalues of at least 100 mm/a. Rapid eastward advance of theorogen into the foreland continued during the late Eocene-Oligocene, as deformation progressed to the Servita fault atrates of 2.2–7.7 mm/a. Taken together, average rates for themiddle Eocene-early Miocene (�40–20 Ma) episode offaster advance of the orogenic deformation front that resultfrom tectonic inversion of the Eastern Cordillera are 4.1–18.0 mm/a. Our data illustrate that by the early Miocene(�20–25 Ma) the northern Andes had achieved �86% oftheir present width. Finally, time-averaged orogen migrationrates slowed since then to 1.2–2.1 mm/a, as the deformationfront remained stationary along the Servita-Tesalia faultsduring the Miocene and propagated to the Guaicaramo faultonly after �5 Ma [Mora, 2007].[42] Our observations demonstrate a nonsystematic

growth of the orogen in the northern Andes characterizedby alternating periods of slow and rapid advance of theorogenic deformation front. Such episodic behavior isdifferent from the growth patterns that would be expectedduring the self-similar growth of an orogenic wedge system[e.g., Dahlen, 1990; DeCelles and DeCelles, 2001]. Inorogenic wedges that deform according to critical Coulombwedge mechanics and under constant accretionary influxand erosive efflux rates, a systematic decrease in the rate ofwedge propagation through time is expected because (1) asthe wedge enlarges, greater amounts of mass influx arerequired by the wedge to continue growing [DeCelles andDeCelles, 2001] and (2) as the wedge propagates into theforeland above a progressively gentler basal decollement, as

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

21 of 27

TC4004

is the case of many orogenic wedges [e.g., Boyer, 1995],greater amounts of convergence are absorbed as internaldeformation in order to build taper, reducing therefore theamount of convergence consumed in orogenic advance[Boyer, 1995]. Alternatively, changes in the rate of orogenicfront propagation may be related to variations in the massflux balance through time [e.g., Dahlen and Barr, 1989].

This model predicts that either an increase in accretionaryinflux rates (for example by an increase in convergencerates) or a decrease in the erosive efflux through timepromote attainment of supercritical conditions that favorfaster orogenic advance [e.g., DeCelles and DeCelles,2001]. Although developed for internally homogeneousorogenic wedges, this concept of mass flux balance as a

Figure 10. Spatial and temporal variation in the position of the deformation front of the northern Andesin central Colombia. (a and b) Present-day and retrodeformed structural configuration along cross sectionA-A0 after Mora et al. [2008] (see location in Figure 1). (c) Rates of orogen propagation. The present(solid symbols) and restored (open symbols) distances of key structures with respect to the undeformedforeland beneath the Guiacaramo thrust are plotted versus their age of onset of deformation. Dotted lineswith arrows indicate structures projected for the analysis (see Table 3). An episodic pattern of Cenozoicorogen front migration shows three major stages. This configuration results from (1) a change fromoblique and slow to orogen-perpendicular and faster convergence in the middle Eocene (�50 Ma) and(2) slip and deformation preferentially accommodated along former normal faults reactivated incontraction. See text for discussion.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

22 of 27

TC4004

major control in orogen width may also apply to those partsof orogens where the crust has been pervasively deformedin previous tectonic episodes and is hence predisposed toconcentrate deformation along basement inhomogenities[Hilley et al., 2005; Streckler et al., 2009], as is the caseof the northern Andes. On the basis of these predictions, theobserved increase in orogenic propagation rates in themiddle Eocene in the central Colombian Andes, �40 Maafter the initiation of orogenesis, should reflect an increasein the net mass flux balance incorporated into the orogen.We suggest that such an increase may have resulted from(1) a middle Eocene (49–42 Ma) increase in the conver-gence rate between the Farallon-Nazca and South Americanplates [Pardo-Casas and Molnar, 1987] and (2) a change inthe convergence direction [Pilger, 1984; Pardo-Casas andMolnar, 1987] that contributed to rotation of the paleostressorientation from a prevailing oblique, WSW-ENE directionprior to the middle Eocene to a subsequent, more orogen-perpendicular NW-SE orientation [Cortes et al., 2005]. As aresult of this change, an increase in the net component ofconvergence accommodated perpendicular to the structuralgrain may have resulted in faster advance of the orogenicdeformation front.[43] On the other hand, a nonsystematic pattern of

orogenic propagation is expected in those settings wherefavorably oriented basement anisotropies within an orogenbehave as weak structures that may favorably absorb short-ening independent of wedge mechanics [e.g., Allmendingeret al., 1983; Jordan and Allmendinger, 1986; Marshak etal., 2000; DeCelles, 2004; Hilley et al., 2005]. Our datasuggest that the extensional structures of the Mesozoicrift basin such as the Soapaga and Servita faults wereutilized and incorporated into the fold-and-thrust beltduring the middle Eocene-late Oligocene contraction.As a result of this predisposition to failure, the rate oforogenic propagation decreased during the Miocene, anddeformation was absorbed along the Servita fault, thuscausing a stagnation of the lateral advance of the orogenicfront. Finally, as shown by Mora [2007], despite the

apparent increase in erosive efflux in the Pliocene docu-mented by young AFT ages, the attainment of elevatedtopography in Eastern Cordillera must have increased thelithostatic stresses in the range and prompted an eastwardjump of the deformation front to low elevation areas, wherecontractional reactivation and failure occurred along theGuaicaramo thrust.

7. Summary and Conclusions

[44] We interpret the thermochronogical ages as reflect-ing erosional exhumation triggered by mountain buildingassociated with shortening. Our main thermochronologicresults are summarized as follows: (1) Reset ZFT ages in thewestern sector of the Eastern Cordillera, along the VilletaAnticlinorium, reveal that exhumation began before 25 Main this sector of the range. (2) Farther to the east, reset AFTages from the Lower Cretaceous strata in the FlorestaMassif and thermal modeling of AFT data from partiallyreset Cenozoic sandstones in the axial Eastern Cordillerasuggest initiation of cooling sometime between 40 and30 Ma. (3) ZFT and AFT ages in the Quetame Massif,along with the structural relations of the contractionallyreactivated faults, suggest onset of exhumation at �20 Ma.(4) Samples from the Quetame Massif suggest that apparentexhumation rates during the Miocene were lower than in thePliocene. Faster exhumation in the Pliocene has beenindependently documented in the Quetame Massif by meansof AFT age-elevation profiles [Mora et al., 2008]. Theconsistency of these two results demonstrates the validityof our approach.[45] Long-term propagation of the northern Andes in

central Colombia occurred at rates of 2.5–2.7 mm/a.However, this advance occurred in three stages. An initialUpper Cretaceous-early Eocene episode of orogen migra-tion occurred at rates of 0.5–3.1 mm/a (1s error) andaccounts for up to 26% of the present width of the orogen.This episode was followed by very rapid orogenic advanceduring the middle Eocene-early Miocene, at average rates of

Table 3. Temporal Evolution of Orogenic Width and Orogenic Migration Rates

FaultAgea

(Ma)1s Error(Ma)

Minimum Maximum

ShortenedDistanceb

(km)

MeanPropagationRatec (mm/a)

1s Error(mm/a)

Percentof OrogenWidth

RestoredDistanced

(km)

MeanPropagationRatee (mm/a)

1s Error(mm/a)

Percent ofOrogenWidth

Guaicaramo 4.0 1.5 0 1.5 0.3 100 0 1.8 0.3 100Servita 22.0 3.0 27 4.0 1.8 86 32 5.3 2.4 87Soapaga 35.0 5.0 79 150 3100 60 101 171 3100 60Salina 35.5 7.5 156 1.6 1.1 21 187 1.8 1.3 26Cambao 49.5 6.5 178 0.7 0.2 10 213 1.4 0.7 16Central Cordillera 77.6 3.9 197 0 252 0

aAge for onset of deformation.bPresent (i.e., shortened) distance to Guaicaramo thrust.cMinimum error-weighted orogen propagation rate (Dshortened distance/Dtime).dRestored distance to Guaicaramo fault according to retrodeformed profile by Mora et al. [2008] for the Eastern Cordillera, and Gomez et al. [2003] for

the Central Cordillera.eMaximum error-weighted orogen propagation rate (Drestored distance/Dtime).

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

23 of 27

TC4004

4.1–18.0 mm/a, which led to the inversion of the EasternCordillera. We propose that this acceleration in the orogenicadvance was produced by an increase in the accretionaryflux that resulted from an increase in the plate convergencerate and a decrease in its obliquity with respect to theinherited structural grain. By the end of this episode of fastpropagation, the orogen had reached �86% of its presentwidth. Finally, slow propagation resulted from stagnation ofthe deformation front along the Servita fault during theMiocene. Preferential accommodation of deformation alongthis crustal anisotropy inherited from the Mesozoic riftingresulted in rates of orogenic advance of 1.2–2.1 mm/a.[46] Our data demonstrate that episodic orogenic advance

may result from contractional reactivation of prestrainedbasement in the orogenic foreland, whereby weak anisotro-pies inherited from previous tectonic events tend to concen-trate slip and deformation. In conclusion, our data setillustrates the need for integrating structural reconstructionsof uplifting mountain ranges with thermochronologic datato provide a comprehensive and more realistic account ofthe complex processes governing orogenic growth.

Appendix A: Analytical Procedures

A1. Apatite Fission Track Analyses

[47] Apatite preparation and analysis were done by Apatiteto Zircon Inc. Apatite and zircon grains were concentratedfollowing conventional heavy liquids and magnetic separa-tion procedures. Apatite grains were then immersed in anepoxy resin and cured at 90�C for 1 h. After grinding andpolishing to expose internal surfaces, the apatites were etchedin 5.5 HNO3 for 20.0 s (±0.5 s) at 21�C (±1�C) to revealspontaneous tracks. The mounts were then scanned to searchfor suitable apatite grains for age dating and those grainlocations digitally recorded. For each suitable grain, repre-sentative kinetic parameters (Dpar) were measured, and thenatural fission track densities were counted. These grainlocalities were then revisited using the LA-ICP-MS; spotanalyses to determine the concentration of U were completedon the identical areas of each grain from which the naturalfission track densities were first counted. The grain mountswere then irradiated with approximately 107 tracks/cm2

fission fragments from a 50 mCi (activity as of July 1996)252Cf source in a vacuum chamber in order to enhance thenumber of confined tracks available for measurement [e.g.,Donelick and Miller, 1991; Donelick et al., 2005]. Irradiatedgrain mounts were then reimmersed in 5.5N HNO3 for 20.0 s(±0.5 s) at 21�C (±1�C) to reveal any horizontal, confinedfission tracks, and the freshly exposed confined tracks werethen measured. Both track lengths and the angle to the c axiswere measured using unpolarized light.[48] For each apatite grain from which fission track age

or length data were collected, an arithmetic mean valuefor the kinetic parameter Dpar [Donelick et al., 2005] wasdetermined from 1 to 4 measurements. Analyses byP. O’Sullivan were completed using 2000� magnification(100� dry objective, 1.25� projection tube, 16� oculars).Natural fission tracks form as a result of the spontaneousnuclear fission of trace amounts of 238U within an apatite or

zircon grain. Using a modified version of the radioactivedecay equation, the fission track age of an apatite or zircongrain can be calculated using the ratio of the number offission tracks intersecting the surface over a unit area in thegrain to the amount of 238U present in the grain. Laserablation inductively coupled plasma-mass spectrometer(LA-ICP-MS) was used to determine the 238U concentra-tions by measuring the ratio of 238U to 43Ca for apatite fromthe exact regions on the individual grains from which thespontaneous tracks were initially counted [see Hasebe et al.,2004; Donelick et al., 2005]. Operating details for theanalyses are listed in Table A1. The fundamental assump-tion is made that Ca occurs in stoichiometric amounts in allapatite grains analyzed by the LA-ICP-MS. The isotope43Ca is used as the indicator of the volume of apatiteablated. Samples are ablated in a helium atmosphere toreduce condensation and elemental fractionation.[49] Spot analyses were used with the laser centered on a

fixed point. A total of 30 scans for 238U, 232Th, 147 Sm, and43Ca (apatite) were performed for each spot analyzed. Ofthese scans, approximately 10 were performed while thelaser was warming up and was blocked from contacting thegrain surface. During this time, background counts werecollected. Once the laser was permitted to hit the grainsurface, a cylindrical pit was excavated to a depth beyond

Table A1. ICP-MS and Laser Operating Conditions and Data

Acquisition Parametersa

Parameters Value/Description

ICP-MS: operatingconditions

Instrument Finnigan Element II MagneticSector ICP-MS

Forward power 1.25 kWReflected power <5 WPlasma gas ArCoolant flow 15 L/minCarrier flow 1.0 L/min (Ar) 0.8 L/min (He),

optimized dailyAuxiliary flow 0.9 L/min

ICP-MS: Aacquisitionparameters

Dwell time 18 ms per peak pointPoints per peak 4Mass window 5%Scans 30Data acquisition time 22 sData acquisition mode electronic scanningIsotopes measured 43Ca (apatite) or 29Si (zircon),

238U, 232Th, and 147 SmLaser: operating conditions

Laser type New Wave UP213 (Nd: YAG)Wavelength 213 nmLaser mode Q switchedLaser output power 8 J/cmLaser warm up time 6 sShot repetition rate 5 HzSampling scheme spot (16 mm apatite, 12 mm zircon)

aAll laser ablation of samples for age analyses were performed under thedirection of Paul B. O’Sullivan at the Washington State University Schoolof Earth and Environmental Sciences GeoAnalytical Laboratory inPullman, Washington.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

24 of 27

TC4004

which uranium did not contribute fission tracks to theetched grain surface. Between 15 and 20 scans performedduring pit excavation were required to reach this depth. Thedepths of a representative number of these pits weremeasured and the 238U/42Ca value for the pit as a wholewas determined on the basis of the weighted mean of the238U/43Ca value for individual scans relative to the depthsfrom which the ablated material was derived.[50] The fission track ages and errors were calculated

using (1) the ratio of the density of natural fission trackspresent in the grain to the amount of 238U present and (2) amodified version of the radioactive decay equation thatincludes a LA-ICP-MS zeta calibration factor (seeequations (1b) for age equation and (2b) for error calcula-tion in the work by Donelick et al. [2005]). This zetacalibration factor is determined for each sample analyzedduring each LA-ICP-MS session by analyzing the U:Ca ratio(for apatite) or U:Si ratio (for zircon) of apatite or zirconcalibration standards with known ages at the beginning andend of each session on the LA-ICP-MS. The standards usedare Durango apatite, 30.6 ± 0.3 Ma, and Fish Canyon zircon,28.5 ± 0.3 Ma. Ages are reported at the 1s level.

A2. Zircon Fission Track Analyses

[51] Zircon tracks analyses were conducted at the Uni-versity of Potsdam. Individual zircon samples were splitinto two to four aliquots and each aliquot etched separatelyin the eutectic mixture of NaOH/KOH at 220�C for periodsbetween 8 and 72 h. Analytical procedures followed theexternal detector method [Gleadow, 1981]. Samples wereirradiated at the Oregon State University research reactor.Following irradiation, mica monitors were etched with 40%hydrofluoric acid at 21�C for 45 min. Samples were countedat a magnification of 1600 � (dry objective) using a LeicaDMRM microscope with drawing tube located above adigitizing tablet, and a Kinetec2 computer-controlled stagedriven by the FTStage program [Dumitru, 1993]. Ages anderrors were calculated using the zeta calibration method[Hurford and Green, 1983] with the software Trackkey

[Dunkl, 2002], and are reported at the 1s level. Zeta calibra-tion factor is 121.8 ± 5.2 for zircon (CN2 glass; M. Parra).

Appendix B: Thermal Modeling

[52] Inverse thermal modeling of AFT data was per-formed with the program HeFTy version 1.5.4 [Ehlers etal., 2005; Ketcham, 2005] using the apatite annealingkinetics model of Ketcham et al. [2007b]. We used LA-ICP-MS-derived age data and c axis corrected track lengths[Ketcham et al., 2007a] for kinetically homogeneous pop-ulations. To further constrain the number of allowed t-Tpaths, burial temperatures derived from vitrinite reflectance(Table S2) according to the correlation model of Burnhamand Sweeney [1989] were introduced into the models.Additional constraints based on the known geologicalhistory of the samples were also introduced in each model(see sections 5.1 and 5.3). Two randomly spaced, mono-tonic consistent half segments (2E) between adjacent con-straints were used. We ran 2,500 to 150,000 iterations usinga Monte Carlo search method until a number between 5(sample A19) and 100 good fits was achieved.

[53] Acknowledgments. We are grateful to Hocol S.A. and Ecopetrolfor permission to publish the apatite fission track data presented in thiswork. P. O’Sullivan from Apatite to Zircon Inc. conducted the unpublishedAFT analyses and provided detailed methodologic information presented inAppendix A. Diane Seward (ETH Zurich) generously provided informationand advice on the zircon fission track methodology and sample preparation.Elias Gomez and Ian Duddy (Geotrack International) kindly provided rawdata from published AFT samples from the Magdalena Basin. We thankBrian Horton, Paolo Ballato, and Carlos Jaramillo for helpful discussionsand Birgit Fabian for graphic work. We also thank A. Velazquez, J. G.Vargas, and Y. Garcıa for discussions and help during the field work. Themanuscript was significantly improved by thorough reviews of PeterDeCelles and an anonymous reviewer. Also, exhaustive revisions of anearlier version of this work by two anonymous reviewers helped to clarifythe ideas discussed in this paper. This study was supported by grants andfellowships from the German Academic Exchange Service (DAAD) to M.Parra and A. Mora, the German Research Foundation (DFG), Str 373/19-1to M. Strecker, funds from the Leibniz Center for Earth Surface and ClimateStudies at Potsdam University, and Universidad Nacional de Colombia(Beca de Honor to M. Parra).

ReferencesAllmendinger, R. W., V. A. Ramos, T. E. Jordan,

M. Palma, and B. L. Isacks (1983), Paleogeogra-phy and Andean structural geometry, northwestArgentina, Tectonics, 2, 1 – 16, doi:10.1029/TC002i001p00001.

Bachu, S., J. C. Ramon, M. E. Villegas, and J. R.Underschultz (1995), Geothermal regime and ther-mal history of the Llanos Basin, Colombia, AAPGBull., 79, 116 –129.

Bayona, G., F. Lamus, A. Cardona, C. Jaramillo,C. Montes, and N. Tchegliakova (2007), Procesosorogenicos para la Cuenca de Rancherıa (Guajira,Colombia) y areas adyacentes definidos por analisisde procedencia, Rev. Geol. Colombiana, 32, 21 –46.

Bayona, G., M. Cortes, C. Jaramillo, G. Ojeda, J. J.Aristizabal, and A. Reyes-Harker (2008), An inte-grated analysis of an orogen-sedimentary basinpair: Latest Cretaceous-Cenozoic evolution of thelinked Eastern Cordillera orogen and the Llanosforeland basin of Colombia, Geol. Soc. Am. Bull.,120, 1171–1197, doi:10.1130/B26187.1.

Bernet, M., and J. I. Garver (2005), Fission-track ana-lysis of detrital zircon, in Low Temperature Ther-

mochronology: Techniques, Interpretations, andApplications, Rev. Mineral. Geochem., vol. 58, edi-ted by P. W. Reiners and T. A. Ehlers, pp. 205–238,Mineral. Soc. of Am., Washington, D. C.

Bertotti, G., P. Mosca, J. Juez, R. Polino, and T. Dunai(2006), Oligocene to present kilometres scale sub-sidence and exhumation of the Ligurian Alps andthe Tertiary Piedmont Basin (NW Italy) revealedby apatite (U-Th)/He thermochronology: Correlationwith regional tectonics, Terra Nova, 18, 18 – 25,doi:10.1111/j.1365-3121.2005.00655.x.

Boyer, S. E. (1995), Sedimentary basin taper as a factorcontrolling the geometry and advance of thrustbelts, Am. J. Sci., 295, 1220–1254.

Brandon, M. T., M. K. Roden-Tice, and J. I. Carver(1998), Late Cenozoic exhumation of the Casca-dia accretionary wedge in the Olympic Moun-tains, northwest Washington State, Geol. Soc.

Am. Bull., 110, 985 – 1009, doi:10.1130/0016-7606(1998)110<0985:LCEOTC>2.3.CO;2.

Bray, R. J., P. F. Green, and I. R. Duddy (1992), Thermalhistory reconstruction using apatite fission trackanalysis and vitrinite reflectance: A case study fromthe UK East Midlands and southern North Sea, inExploration Britain: Geological Insights for theNext Decade, edited by R.S.P. Hardman, Geol.Soc. Spec. Publ., 67, 3 – 25.

Burke, K. (1988), Tectonic evolution of the Caribbean,Annu. Rev. Earth Planet. Sci., 16, 201 – 230,doi:10.1146/annurev.ea.16.050188.001221.

Burnham, A. K., and J. J. Sweeney (1989), A chemicalkinetic model of vitrinite maturation and reflectance,Geochim. Cosmochim. Acta, 53, 2649 – 2657,doi:10.1016/0016-7037(89)90136-1.

Cespedes, S., and L. Pena (1995), Relaciones estrati-graficas y ambientes de deposito de las forma-ciones del Terciario Inferior aflorante entre Tunjay Paz de Rıo, Boyaca, 50 pp., Univ. Nac. deColombia, Bogota.

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

25 of 27

TC4004

Colmenares, L., and M. D. Zoback (2003), Stress fieldand seismotectonics of northern South America,Geology, 31, 721–724, doi:10.1130/G19409.1.

Coney, P. J., and C. A. Evenchick (1994), Consolida-tion of the American Cordilleras, J. South Am.Earth Sci., 7, 241 – 261, doi:10.1016/0895-9811(94)90011-6.

Cooper, M. A., et al. (1995), Basin development andtectonic history of the Llanos Basin, Eastern Cor-dillera, and Middle Magdalena Valley, Colombia,AAPG Bull., 79, 1421–1443.

Cortes, M., and J. Angelier (2005), Current states ofstress in the northern Andes as indicated by focalmechanisms of earthquakes, Tectonophysics, 403,29– 58, doi:10.1016/j.tecto.2005.03.020.

Cortes, M., J. Angelier, and B. Colleta (2005), Paleos-tress evolution of the northern Andes (Eastern Cor-dillera of Colombia): Implications on platekinematics of the South Caribbean region, Tec-tonics, 24, TC1008, doi:10.1029/2003TC001551.

Cortes,M., B. Colletta, and J. Angelier (2006), Structureand tectonics of the central segment of the EasternCordillera of Colombia, J. South Am. Earth Sci., 21,437–465, doi:10.1016/j.jsames.2006.07.004.

Crowhurst, P. V., P. F. Green, and P. J. J. Kamp (2002),Appraisal of (U-Th)/He apatite thermochronologyas a thermal history tool for hydrocarbon explora-tion: An example from the Taranaki Basin, NewZealand, AAPG Bull., 86, 1801–1819.

Dahlen, F. A. (1990), Critical taper model of fold-and-thrust belts and accretionary wedges, Annu. Rev.Earth Planet. Sci., 18, 55– 99, doi:10.1146/annur-ev.ea.18.050190.000415.

Dahlen, F. A., and T. D. Barr (1989), Brittle frictionalmountain building: 1. Deformation and mechanicalenergy budget, J. Geophys. Res., 94, 3906–3922,doi:10.1029/JB094iB04p03906.

Dahlen, F. A., J. Suppe, and D. Davis (1984),Mechanicsof fold-and-thrust belts and accretionary wedges:Cohesive Coulomb theory, J. Geophys. Res., 89,10,087–10,101, doi:10.1029/JB089iB12p10087.

Davis, D., J. Suppe, and F. A. Dahlen (1983), Me-chanics of fold-and-thrust belts and accretionarywedges, J. Geophys. Res., 88, 1153 – 1172,doi:10.1029/JB088iB02p01153.

DeCelles, P. G. (2004), Late Jurassic to Eocene evolu-tion of the Cordilleran thrust belt and foreland basinsystem, western USA, Am. J. Sci., 304, 105–168,doi:10.2475/ajs.304.2.105.

DeCelles, P. G., and P. C. DeCelles (2001), Rates ofshortening, propagation, underthrusting, and flexur-al wave migration in continental orogenic systems,Geology, 29 , 135 – 138, doi:10.1130/0091-7613(2001)029<0135:ROSPUA>2.0.CO;2.

DeCelles, P. G., and G. Mitra (1995), History of theSevier orogenic wedge in terms of critical tapermodels, northeast Utah and southwest Wyoming,Geol. Soc. Am. Bull., 107, 454–462, doi:10.1130/0016-7606(1995)107<0454:HOTSOW>2.3.CO;2.

Dengo, C. A., and M. C. Covey (1993), Structure ofthe Eastern Cordillera of Colombia: Implicationsfor trap styles and regional tectonics, AAPG Bull.,77(8), 1315–1337.

Dimate, C., L. A. Rivera, A. Taboada, B. Delouis, A.Osorio, E. Jimenez, A. Fuenzalida, A. Cisternas,and I. Gomez (2003), The 19 January 1995 Taur-amena (Colombia) earthquake: Geometry andstress regime, Tectonophysics, 363, 159 – 180,doi:10.1016/S0040-1951(02)00670-4.

Donelick, R., and D. S. Miller (1991), Enhanced TINTfission track densities in low spontaneous trackdensity apatites using 252Cf-derived fission frag-ment tracks. A model and experimental observa-tions, Nucl. Tracks Radiat. Meas., 18, 301–307,doi:10.1016/1359-0189(91)90022-A.

Donelick, R. A., R. A. Ketcham, and W. D. Carlson(1999), Variability of apatite fission-track anneal-ing kinetics: II. Crystallographic orientation effects,Am. Mineral., 84, 1224–1234.

Donelick, R. A., P. B. O’Sullivan, and R. A. Ketcham(2005), Apatite fission-track analysis, in Low Tem-

perature Thermochronology: Techniques, Interpre-

tations, and Applications, Rev. Mineral. Geochem.,vol. 58, edited by P. W. Reiners and T. A. Ehlers,pp. 49–94,Mineral. Soc. of Am.,Washington, D. C.

Dumitru, T. A. (1993), New computer-automated mi-croscope stage system for fission-track analysis,Nucl. Tracks Radiat. Meas., 21, 575 – 580,doi:10.1016/1359-0189(93)90198-I.

Dunkl, I. (2002), Trackkey: A windows program forcalculation and graphical presentation of fissiont rack da ta , Comput . Geosc i . , 28 , 3 – 12,doi:10.1016/S0098-3004(01)00024-3.

Echavarrıa, L., R. Hernandez, R. Allmendinger, and J.Reynolds (2003), Subandean thrust and fold belt ofnorthwestern Argentina: Geometry and timing ofthe Andean evolution, AAPG Bull., 87, 965–985,doi:10.1306/01200300196.

Ehlers, T. A. (2005), Crustal thermal processes and theinterpretation of thermochronometer data, in Low

Temperature Thermochronology: Techniques, In-terpretations, and Applications, Rev. Mineral. Geo-chem., vol. 58, edited by P. W. Reiners and T. A.Ehlers, pp. 315 – 350, Mineral. Soc. of Am.,Washington, D. C.

Ehlers, T. A., et al. (2005), Computational tools forlow-temperature thermochronometer interpretation,in Low Temperature Thermochronology: Techni-ques, Interpretations, and Applications, Rev.Mineral. Geochem., vol. 58, edited by P. W. Rein-ers and T. A. Ehlers, pp. 589 –622, Mineral. Soc.of Am., Washington, D. C.

Erslev, E. A. (1986), Basement balancing of RockyMountain foreland uplifts (USA), Geology, 14,259 – 262, doi:10.1130/0091-7613(1986)14<259:BBORMF>2.0.CO;2.

Etayo-Serna, F. (1968), El sistema Cretaceo en la re-gion de Villa de Leiva y zonas proximas, Rev.Geol. Colombiana, 5, 5 –74.

Fitzgerald, P. G., E. Stump, and T. F. Redfield (1993),Late Cenozoic uplift of Denali and its relation torelative plate motion and fault morphology,Science, 259, 497 – 499, doi:10.1126/science.259.5094.497.

Galbraith, R. F. (1981), On statistical models for fis-sion-tracks counts, Math. Geol., 13, 471 – 478,doi:10.1007/BF01034498.

Garver, J. I., P. W. Reiners, L. J. Walker, J. M. Ramage,and S. E. Perry (2005), Implications for timing ofAndean uplift from thermal resetting of radiation-damaged zircon in the Cordillera Huayhuash,nor thern Peru , J . Geol . , 113 , 117 – 138,doi:10.1086/427664.

Gleadow, A. J. W. (1981), Fission-track dating meth-ods: What are the real alternatives?, Nucl. Tracks,5, 3 –14, doi:10.1016/0191-278X(81)90021-4.

Godin, L., R. R. Parrish, R. L. Brown, and K. V.Hodges (2001), Crustal thickening leading to ex-humation of the Himalayan metamorphic core ofCentral Nepal: Insight from U-Pb geochronologyand 40Ar/39Ar thermochronology, Tectonics, 20,729–747, doi:10.1029/2000TC001204.

Gomez, A., C. Jaramillo, M. Parra, and A. Mora(2009), Huesser Horizon: A lake and marine incur-sion in northwestern South America during theearly Miocene, Palaios , 24(4), 199 – 210,doi:10.2110/palo.2007.p07-074r.

Gomez, E. (2001), Tectonic controls on the Late Cre-taceous to Cenozoic sedimentary fill of the MiddleMagdalena Valley Basin, Eastern Cordillera andLlanos Basin, Colombia, Ph.D. thesis, 619 pp.,Cornell Univ., Ithaca, N. Y.

Gomez, E., T. E. Jordan, R.W. Allmendinger, K. Hegarty,S. Kelley, and M. Heizler (2003), Controls on archi-tecture of the Late Cretaceous to Cenozoic southernMiddle Magdalena Valley Basin, Colombia, Geol.Soc. Am. Bull., 115, 131–147, doi:10.1130/0016-7606(2003)115<0131:COAOTL>2.0.CO;2.

Gomez, E., T. E. Jordan, R. W. Allmendinger, and N.Cardozo (2005), Development of the Colombianforeland-basin system as a consequence of diachro-nous exhumation of the northern Andes, Geol. Soc.Am. Bull. , 117 , 1272 – 1292, doi:10.1130/B25456.1.

Gomez, J., et al. (2007), Geologic map of Colombia,INGEOMINAS, Bogota.

Green, P. F. (1981), A new look at statistics in fission-track dating, Nucl. Tracks, 5, 77–86, doi:10.1016/0191-278X(81)90029-9.

Green, P. F., I. R. Duddy, A. J. W. Gleadow, P. R.Tingate, and G. M. Laslett (1986), Thermal anneal-ing of fission tracks in apatite 1. A qualitative de-scription, Chem. Geol., 59, 237–253, doi:10.1016/0009-2541(86)90048-3.

Hasebe, N., J. Barbarand, K. Jarvis, A. Carter, and A. J.Hurford (2004), Apatite fission-track chronometryusing laser ablation ICP-MS, Chem. Geol., 207,135 –145, doi:10.1016/j.chemgeo.2004.01.007.

Hilley, G. E., P. M. Blisniuk, and M. R. Strecker(2005), Mechanics and erosion of basement coreduplift provinces, J. Geophys. Res., 110, B12409,doi:10.1029/2005JB003704.

Hoth, S., J. Adam, N. Kukowski, and O. Onken(2005), Influence of erosion on the kinematics ofbivergent orogens: Results from scaled sandboxsimulations, in Tectonics, Climate, and LandscapeEvolution, edited by S. D. Willet et al., Spec. Pap.Geol. Soc. Am., 398, 201–225.

Hubbard, M. S., and T. M. Harrison (1989), 40Ar/39Arage constraints on deformation and metamorphismin the Main Central Thrust zone and Tibetan Slab,eastern Nepal, Himalaya, Tectonics, 8, 865–880,doi:10.1029/TC008i004p00865.

Hurford, A. J., and P. F. Green (1983), The zeta agecalibration of fission-track dating, Isot. Geosci., 1,285 –317.

Jones, M. A., P. L. Heller, E. Roca, M. Garces, andL. Cabrera (2004), Time lag of syntectonic sedimen-tation across an alluvial basin: Theory and examplefrom the Ebro Basin, Spain, Basin Res., 16, 489–506, doi:10.1111/j.1365-2117.2004.00244.x.

Jordan, T. E. (1981), Thrust loads and foreland basinevolution, Cretaceous, western United States,AAPG Bull., 65, 2506–2520.

Jordan, T. E., and R. W. Allmendinger (1986), TheSierras Pampeanas of Argentina: A modern analo-gue of Rocky Mountain foreland deformation, Am.J. Sci., 286, 737–764.

Julivert, M. (1963), Los rasgos tectonicos de la regionde la Sabana de Bogota y los mecanismos de laformacion de las estructuras, Bol. Geol. Univ. Ind.Santander, 13–14, 5 –102.

Kammer, A., and J. Sanchez (2006), Early Jurassic riftstructures associated with the Soapaga and Boyacafaults of the Eastern Cordillera, Colombia: Sedi-mentological inferences and regional implications,J . Sou th Am. Ear th Sc i . , 21 , 412 – 422,doi:10.1016/j.jsames.2006.07.006.

Kerr, A. C., and J. Tarney (2005), Tectonic evolution ofthe Caribbean and northwestern South America:The case for accretion of two Late Cretaceous ocea-nic plateaus, Geology, 33, 269–272, doi:10.1130/G21109.1.

Ketcham, R. A. (2005), Forward and inverse modelingof low-temperature thermochronometry data, inLow Temperature Thermochronology: Techniques,Interpretations, and Applications, Rev. Mineral.Geochem., vol. 58, edited by P. W. Reiners andT. A. Ehlers, pp. 275–314, Mineral. Soc. of Am.,Washington, D. C.

Ketcham, R. A., R. A. Donelick, and W. D. Carlson(1999), Variability of apatite fission-track anneal-ing kinetics: III. Extrapolation to geological timescales, Am. Mineral., 84, 1235–1255.

Ketcham, R. A., A. Carter, R. A. Donelick, J. Barbarand,and A. J. Hurford (2007a), Improved measurementof fission-track annealing in apatite using c-axisprojection, Am. Mineral. , 92 , 789 – 798,doi:10.2138/am.2007.2280.

Ketcham, R. A., A. Carter, R. A. Donelick, J. Barbarand,and A. J. Hurford (2007b), Improved modeling offission-track annealing in apatite, Am. Mineral., 92,799 –810, doi:10.2138/am.2007.2281.

Mancktelow, N. S., and B. Grasemann (1997), Time-dependent effects of heat advection and topographyon cooling histories during erosion, Tectonophy-

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

26 of 27

TC4004

s ics , 270 , 167 – 195, doi :10.1016/S0040-1951(96)00279-X.

Marshak, S., K. Karlstrom, and J. M. Timmons (2000),Inversion of Proterozoic extensional faults: An ex-planation for the pattern of Laramide and AncestralRockies intracratonic deformation, United States,Geology, 28, 735 – 738, doi:10.1130/0091-7613(2000)28<735:IOPEFA>2.0.CO;2.

Mayorga, M., and M. Vargas (1995), Caracterizaciongeoquımica y facial de las rocas potencialmentegeneradoras de hidrocarburos del Cretaceo y Terciarioinferior de la Cordillera Oriental y Piedemomte,B.Sc. thesis, 150 pp., Univ. Nac. de Colombia,Bogota.

McCourt, W. J., J. A. Aspden, and M. Brook (1984),New geological and geochronological data fromthe Colombian Andes: Continental growth by mul-tiple accretion, J. Geol. Soc., 141, 831 – 845,doi:10.1144/gsjgs.141.5.0831.

Mojica, J., and C. Villarroel (1984), Contribucion alconocimiento de las unidades paleozoicas del areade Floresta (Cordillera Oriental Colombiana, De-partamento de Boyaca) y en especial de la Forma-cion Cuche, Rev. Geol. Colombiana, 13, 55– 80.

Montes, C., R. D. Hatcher, Jr., and P. A. Restrepo-Pace(2005), Tectonic reconstruction of the northern An-dean blocks: Oblique convergence and rotationsderived from the kinematics of the Piedras-Girardotarea, Colombia, Tectonophysics, 399, 221 – 250,doi:10.1016/j.tecto.2004.12.024.

Mora, A. (2007), Inversion tectonics and exhumationprocesses in the Eastern Cordillera of Colombia,Ph.D. thesis, 133 pp., Univ. Potsdam, Potsdam.

Mora, A., M. Parra, M. R. Strecker, A. Kammer,C. Dimate, and F. Rodriguez (2006), Cenozoiccontractional reactivation of Mesozoic exten-sional structures in the Eastern Cordillera ofColombia, Tectonics, 25, TC2010, doi:10.1029/2005TC001854.

Mora, A., M. Parra, M. R. Strecker, E. R. Sobel,H. Hooghiemstra, V. Torres, and J. Vallejo-Jaramillo(2008), Climatic forcing of asymmetric orogenicevolution in the Eastern Cordillera of Colombia,Geol. Soc. Am. Bull., 120, 930–949, doi:10.1130/B26186.1.

Mora, A., T. Gaona, J. Kley, D. Montoya, M. Parra,L. I. Quiroz, G. Reyes, and M. R. Strecker(2009), The role of inherited extensional fault seg-mentation and linkage in contractional orogenesis:A reconstruction of Lower Cretaceous inverted riftbasins in the Eastern Cordillera of Colombia, BasinRes., 21, 111–137.

O’Sullivan, P. B. (1999), Thermochronology, denuda-tion and variations in palaeosurface temperature: Acase study from the North Slope foreland basin,Alaska, Basin Res., 11, 191 – 204, doi:10.1046/j.1365-2117.1999.00094.x.

O’Sullivan, P. B., and W. K. Wallace (2002), Out-of-sequence, basement-involved structures in theSadlerochit Mountains region of the Arctic NationalWildlife Refuge, Alaska: Evidence and implicationsfrom fission-track thermochronology, Geol. Soc.Am. Bull., 114, 1356 – 1378, doi:10.1130/0016-7606(2002)114<1356:OOSBIS>2.0.CO;2.

Pardo-Casas, F., and P. Molnar (1987), Relative motionof the Nazca (Farallon) and South American platessince Late Cretaceous time, Tectonics, 6, 233–248,doi:10.1029/TC006i003p00233.

Parra, M., A. Mora, C. Jaramillo, M. R. Strecker, andE. R. Sobel (2007), Cenozoic exhumation historyin the northeastern Andes: New data based on low-T thermochronology and basin analysis in the East-ern Cordillera of Colombia, Geophys. Res. Abstr.,9, 07197.

Parra, M., A. Mora, C. Jaramillo, M. R. Strecker, E. R.Sobel, L. I. Quiroz, M. Rueda, and V. Torres(2009), Orogenic wedge advance in the northernAndes: Evidence from the Oligocene-Miocene se-dimentary record of the Medina Basin, Eastern

Cordillera, Colombia, Geol. Soc. Am. Bull., 121,780–800, doi:10.1130/B26257.1.

Pfiffner, O. A., S. Ellis, and C. Beaumont (2000), Col-lision tectonics in the Swiss Alps: Insight fro geo-dynamic modeling, Tectonics, 19, 1065 – 1094,doi:10.1029/2000TC900019.

Pilger, R. H. (1984), Cenozoic plate kinematics, sub-duction and magmatism: South American Andes,J. Geol. Soc., 141, 793 – 802, doi:10.1144/gsjgs.141.5.0793.

Rahn, M. K., M. T. Brandon, G. E. Batt, and J. I. Garver(2004), A zero-damage model for fission-track an-nealing in zircon, Am. Mineral., 89, 473–484.

Ramon, J. C., and A. Rosero (2006), Multiphase struc-tural evolution of the western margin of the Girar-dot subbasin, Upper Magdalena Valley, Colombia,J . South Am. Ear th Sc i . , 21 , 493 – 509 ,doi:10.1016/j.jsames.2006.07.012.

Reiners, P. W., and M. T. Brandon (2006), Using ther-mochronology to understand orogenic erosion,Annu. Rev. Earth Planet. Sci., 34, 419 – 466,doi:10.1146/annurev.earth.34.031405.125202.

Reiners, P. W., Z. Zhou, T. A. Ehlers, C. Xu, M. T.Brandon, R. A. Donelick, and S. Nicolescu (2003),Post-orogenic evolution of the Dabie-Shan, easternChina, from (U-Th)/He and fission-track thermo-chronology, Am. J . Sci . , 303 , 489 – 518,doi:10.2475/ajs.303.6.489.

Restrepo-Pace, P. A., F. Colmenares, C. Higuera, andM. Mayorga (2004), A fold-and-thrust belt alongthe western flank of the Eastern Cordillera of Co-lombia. Style, kinematics, and timing constraintsderived from seismic data and detailed surfacemapping, in Thrust Tectonics and HydrocarbonSystems, edited by K. R. McClay, AAPG Mem.,82, 598–613.

Richardson,N. J., A. L.Densmore, D. Seward, A. Fowler,M.Wipf, M. A. Ellis, L. Yong, and Y. Zhang (2008),Extraordinary denudation in the Sichuan Basin: In-sights from low-temperature thermochronology ad-jacent to the eastern margin of the TibetanPla teau, J . Geophys . Res . , 113 , B04409,doi:10.1029/2006JB004739.

Ring, U., M. T. Brandon, S. D. Willett, and G. S. Lister(1999), Exhumation processes, in Exhumation Pro-cesses; Normal Faulting, Ductile Flow and Ero-

sion, edited by U. Ring et al., Geol. Soc. Spec.Publ., 154, 1 – 27.

Roeder, D., and R. L. Chamberlain (1995), EasternCordillera of Colombia: Jurassic-Neogene crustalevolution, in Petroleum Basins of South America,edited by A. J. Tankard, S. R. Suarez, and H. J.Welsink, AAPG Mem., 62, 633–645.

Russo, R. M., and P. G. Silver (1996), Cordillera for-mation, mantle dynamics, and the Wilson cycle,Geology, 24 , 511 – 514, doi:10.1130/0091-7613(1996)024<0511:CFMDAT>2.3.CO;2.

Sarmiento-Rojas, L. F., J. D. Van Wess, and S. Cloe-tingh (2006), Mesozoic transtensional basin historyof the Eastern Cordillera, Colombian Andes: Infer-ences from tectonic models, J. South Am. EarthSci. , 21 , 383 – 411, doi:10.1016/j . jsames.2006.07.003.

Segovia, A. (1965),MapaGeologico de la plancha L-12

(Medina) de la Republica de Colombia, Serv. Geol.Nac., Bogota, Colombia.

Shagam, R., B. P. Kohn, P. O. Banks, L. E. Dasch,R. Vargas, G. I. Rodrıguez, and N. Pimentel(Eds.) (1984), Tectonic implications of Cretaceous-Pliocene fission-track ages from rocks of theCircum-Maracaibo Basin Region of western Vene-

zuela and eastern Colombia, Mem. Geol. Soc. Am.,162, 385–412.

Sobel, E. R., and M. R. Strecker (2003), Uplift, exhu-mation and precipitation: Tectonic and climaticcontrol of Late Cenozoic landscape evolution inthe northern Sierras Pampeanas, Argentina, BasinRes., 15, 431 – 451, doi:10.1046/j.1365-2117.2003.00214.x.

Spikings, R. A., W.Winkler, D. Seward, and R. Handler(2001), Along-strike variations in the thermal andtectonic response of the continental Ecuadorian An-des to the collision with heterogeneous oceaniccrust, Earth Planet. Sci. Lett., 186, 57 – 73,doi:10.1016/S0012-821X(01)00225-4.

Strecker, M. R., R. Alonso, B. Bookhagen, B. Carrapa,I. Coutand, M. P. Hain, G. E. Hilley, E. Mortimer,L. Schoenbohm, and E. R. Sobel (2009), Does thetopographic distribution of the central AndeanPuna Plateau result from climatic or geodynamicprocesses?, Geology, in press.

Steinmann, M., D. Hungerbuhler, D. Seward, andW. Winkler (1999), Neogene tectonic evolutionand exhumation of the southern Ecuadorian Andes:A combined stratigraphy and fission-track ap-p roach , Tec tonophys i c s , 307 , 255 – 276 ,doi:10.1016/S0040-1951(99)00100-6.

Stockli, D. F. (2005), Application of low-temperaturethermochronometry to extensional tectonic set-tings, in Low Temperature Thermochronology:

Techniques, Interpretations, and Applications,Rev. Mineral. Geochem., vol. 58, edited by P. W.Reiners and T. A. Ehlers, pp. 411–448, Mineral.Soc. of Am., Washington, D. C.

Stockli, D. F., B. E. Surpless, T. A. Dumitru, and K. A.Farley (2002), Thermochronological constraints onthe timing and magnitude of Miocene and Plioceneextension in the central Wassuk Range, westernNevada, Tectonics, 21(4), 1028, doi:10.1029/2001TC001295.

Taboada, A., L. A. Rivera, A. Fuenzalida, A. Cisternas,H. Philip, H. Bijwaard, J. Olaya, and C. Rivera(2000), Geodynamics of the northern Andes: Sub-ductions and intracontinental deformation (Colom-bia), Tectonics, 19, 787 – 813, doi:10.1029/2000TC900004.

Tagami, T., and P. B. O’Sullivan (2005), Fundamentalsof fission-track thermochronology, in Low Tem-

perature Thermochronology: Techniques, Interpre-tations, and Applications, Rev. Mineral. Geochem.,vol. 58, edited by P. W. Reiners and T. A. Ehlers,pp. 19–47, Mineral. Soc. of Am., Washington,D. C.

Tagami, T., and C. Shimada (1996), Natural long-termannealing of the fission track annealing of the fis-sion-track system around a granitic pluton, J. Geo-phys. Res., 101, 8245 – 8255, doi:10.1029/95JB02885.

Toro, J. (1990), The termination of the Bucaramangafault in the Cordillera Oriental, Colombia, M.Sc.thesis, 53 pp., Univ. of Ariz., Tucson.

Toro, J., F. Roure, N. Bordas-Le Floch, S. Le Cornec-Lance, and W. Sassi (2004), Thermal and kinematicevolution of the Eastern Cordillera fold-and-thrust-belt, Colombia, in Deformation, Fluid Flow, andReservoir Appraisal in Foreland Fold and Thrust

Belts, edited by R. Swennen et al., pp. 79 –115,Am. Assoc. of Pet. Geol., Tulsa, Okla.

Ulloa, C., and E. Rodrıguez (1979), Geologıa del Cua-drangulo K12, Guateque, Bol. Geol. Ing. Bogota,22, 3 –55.

Vallejo, C., R. A. Spikings, L. Luzieux, W. Winkler,D. Chew, and L. Page (2006), The early interactionbetween the Caribbean Plateau and the NW SouthAmerican Plate, Terra Nova, 18, 264–269.

���������R. Gonzalez, Hocol S.A., Cra 7 113-43 Piso 16,

Edificio Samsung, Bogota, Colombia.

A. Mora, Instituto Colombiano del Petroleo,Ecopetrol, AA 4185, Bucaramanga, Colombia.

M. Parra, E. R. Sobel, and M. R. Strecker, Institutfur Geowissenschaften, Universitat Potsdam, Karl-Liebknecht-Strasse 24, Haus 27, D-14476 Golm,Germany. (mauricio@geo.uni-potsdam.de)

TC4004 PARRA ET AL.: OROGEN-FRONT MIGRATION, COLOMBIAN ANDES

27 of 27

TC4004